Stationary Diesel Engines in the Northeast:
An Initial Assessment of the Regional Population,
Control Technology Options and Air Quality Policy Issues
Northeast States for Coordinated Air Use Management
101 Merrimac Street, 10th Floor
Boston, MA 02114
(617) 367-8540
www.nescaum.org
June 2003
Founded in 1967, NESCAUM is a non-profit association of the state air quality
management offices of Connecticut, Maine, Massachusetts, New Hampshire,
New Jersey, New York, Rhode Island and Vermont.
The cover photo shows a 315 kW stationary diesel generator, courtesy of:
www.utilitywarehouse.com
Stationary Diesel Engines in the Northeast:
An Initial Assessment of the Regional Population,
Control Technology Options and Air Quality Policy Issues
Project Managers
Carrie Pistenmaa
Praveen Amar
T.J. Roskelley
Report Editor
Marika Tatsutani
Principal Contributors
Executive Summary
Chapter I
Chapter II
Chapters III, IV
Chapter V
Chapters VI, VII
Chapter VIII
Marika Tatsutani, Carrie Pistenmaa, Praveen Amar
Marika Tatsutani
Marika Tatsutani, Carrie Pistenmaa
Carrie Pistenmaa, Marika Tatsutani
Carrie Pistenmaa, Marika Tatsutani, Phil Johnson, Dave Brown
Dale L. McKinnon, Antonio Santos, William Gillespie
(ESI International, Inc.), Praveen Amar
Marika Tatsutani, Carrie Pistenmaa, Praveen Amar
Principal Consultants
ESI International, Inc.
1660 L Street, NW, Suite 1100
Washington, D.C. 20036
(202) 775-8868
Power Systems Research
1365 Corporate Center Curve, 2nd Floor
St. Paul, MN 55121-1298
(651) 905-8487
i
ii
Acknowledgements
We gratefully acknowledge funding support for this project from several offices of the
United States Environmental Protection Agency (EPA), including:
Office of Air and Radiation (OAR)
Clean Air Markets Division, Office of Atmospheric Programs (OAP)
Ozone Policy and Strategies Group, Air Quality Strategies and Standards
Division, Office of Air Quality Planning and Standards (OAQPS)
Integrated Policy and Strategies Group, Air Quality Strategies and Standards
Division, OAQPS
EPA Region I, Boston
Additional funding was provided by the New York Department of Environmental
Conservation and the New Hampshire Department of Environmental Services.
We extend our special appreciation to William White (EPA Region I), William Baker
(EPA Region II) and Bill Neuffer (EPA OAQPS) for their thoughtful guidance
throughout all phases of this effort. We also wish to thank members of the NESCAUM
Stationary Source and Permits Committee (Co-Chairs: Gary Rose, Connecticut and
Randy Orr, New York) for their assistance in designing the scope of the project and for
providing considerable insight concerning state permitting programs.
Finally, we thank numerous individuals at EPA and in state offices for providing valuable
suggestions and assistance over the two-year course of this project. The NESCAUM
Stationary Source and Permits Committee members and other individuals who provided
technical assistance, compiled permit data and reviewed drafts of this report are listed
below.
Connecticut Department of Environmental Protection
Ernest Bouffard, Henry Hampton, Chris James, Gary Rose, William Simpson
Maine Department of Environmental Protection
Marc Cone, Mark Roberts
Massachusetts Department of Environmental Protection
Robert Donaldson, Nancy Seidman, Donald Squires
New Hampshire Department of Environmental Services
Andrew Bodnarik, Joe Fontaine, Sonny Strickland
New Jersey Department of Environmental Protection
Iclal Atay, Yogesh Doshi, William Kuehne, Shirley Meisner, William O’Sullivan,
Viorica Petriman, Francis Steitz, Vidhu Sukheja
New York Department of Environmental Conservation
John Barnes, Michael Jennings, Randy Orr, Robert Sliwinski
iii
New York City Department of Environmental Protection
Geraldine Kelpin
Rhode Island Department of Environmental Management
Douglas McVay
Vermont Department of Environmental Conservation
Douglas Elliot
U.S. Environmental Protection Agency
Joe Bryson (Climate Policy and Programs Division, OAP), Chad Whiteman
(Clean Air Markets Division, OAP), Anthony Gardella (Region II), Edward Linky
(Region II), Bruce Smith (Region III)
We would also like to recognize the effort extended by our consultants at both ESI
International, Inc., and Power Systems Research. These individuals include:
Dale McKinnon, ESI International, Inc.
William Gillespie, (formerly) ESI International, Inc.
Antonio Santos, ESI International, Inc.
George Zirnhelt, Power Systems Research
iv
Table of Contents
Acknowledgements .......................................................................................................... iii
Table of Contents ............................................................................................................. v
List of Tables and Figures .............................................................................................. ix
List of Acronyms and Abbreviations ........................................................................... xiii
Executive Summary ................................................................................................... ES-1
I.
Introduction........................................................................................................... 1
II.
Background and Context ..................................................................................... 5
A. Current Role of Electricity-Generating Diesel Engines in the Northeast ......... 5
B. Summary of Current State Permitting Requirements for Diesel IC Engines.... 7
1. Connecticut ................................................................................................. 8
2. Maine .......................................................................................................... 9
3. Massachusetts ........................................................................................... 10
4. New Hampshire ........................................................................................ 10
5. New Jersey ................................................................................................ 10
6. New York.................................................................................................. 11
7. Rhode Island ............................................................................................. 12
8. Vermont .................................................................................................... 12
III.
Northeast States Inventory................................................................................. 13
A. Introduction..................................................................................................... 13
B. Methodology ................................................................................................... 13
1. PSR Population Estimates........................................................................ 13
2. State Permit Data ....................................................................................... 15
C. Northeast Distributed Generation Inventory Results...................................... 15
1. NESCAUM Region Totals ....................................................................... 15
2. Connecticut Inventory............................................................................... 18
3. Maine Inventory........................................................................................ 20
4. Massachusetts Inventory........................................................................... 21
5. New Hampshire Inventory........................................................................ 23
6. New Jersey Inventory ............................................................................... 24
7. New York Inventory ................................................................................. 26
8. Rhode Island Inventory............................................................................. 27
9. Vermont Inventory.................................................................................... 28
v
IV.
Engine Population Surveys – New York City and Fairfield County, CT ...... 31
A. Introduction..................................................................................................... 31
B. Methodology ................................................................................................... 31
C. Telephone Survey Results............................................................................... 32
1. New York City.......................................................................................... 32
2. Fairfield County, CT................................................................................. 34
D. Overlap between Survey Results and State Permitting Records .................... 35
1. New York City.......................................................................................... 35
2. Fairfield County, Connecticut................................................................... 36
3. Analysis..................................................................................................... 37
V.
Emissions Estimates............................................................................................ 39
A. Health Risks Associated with Diesel Exhaust ................................................ 40
B. Emissions Analysis for New York City and Fairfield County, CT Using PSR
Telephone Survey Results..................................................................................... 42
1. Methodology for Emissions Impact Analysis........................................... 42
2. Estimated Emissions Impact for New York City...................................... 44
3. Estimated Emissions Impact for Fairfield County, Connecticut .............. 44
4. Estimated Summertime Emissions Impact ............................................... 45
C. Emissions Estimates for ISO-Sponsored Demand Response Programs ......... 46
1. New York ISO Program............................................................................ 46
2. New England ISO Programs..................................................................... 47
D. Potential Impact of Real-Time Pricing on Future Emissions from On–Site
Generators ............................................................................................................. 51
VI.
Emission Control Technologies for Stationary Diesel Generators................. 55
A. Distributed Generation Market Status and Outlook........................................ 56
1. National Market ........................................................................................ 57
2. Small Commercial and Residential Markets............................................. 57
3. Combined Heat & Power Market ............................................................. 58
4. Utility Market............................................................................................ 59
B. Emissions Factors for Stationary IC Engines.................................................. 60
C. Emission Control Technologies for Diesel and Natural Gas Stationary IC
Engines Used in Distributed Power Generation ................................................... 60
1. Diesel Particulate Filters (DPFs) .............................................................. 63
2. Diesel Oxidation Catalysts (DOCs) .......................................................... 68
3. Crankcase Emission Controls ................................................................... 71
4. Selective Catalytic Reduction (SCR)........................................................ 72
5. Diesel Particulate Filters Combined with SCR......................................... 73
6. Exhaust Gas Recirculation (EGR) ............................................................ 74
7. Other NOx Control Technologies and Strategies for Lean-Burn Engines 75
vi
8. Ultra-Low Sulfur Fuel as a Control Option for Particulate Matter........... 76
VII.
Control Technology Case Studies...................................................................... 77
A. Cost Methodology........................................................................................... 78
B. Case Study Assumptions ................................................................................ 79
C. Case Study Descriptions.................................................................................. 80
1. Kings County, Department of Public Works, CA..................................... 80
2. National Steel and Shipbuilding Company (NASSCO) ........................... 84
3. Pacific Bell-SBC Telecommunications Facility, San Francisco, CA ....... 87
4. Pacific Bell-SBC Telecommunications Facility, San Jose, CA................ 90
5. Santa Clara County Building Operations.................................................. 93
6. Sierra Nevada Brewing Company, Chico, CA ......................................... 96
D. Cost-Effectiveness Analysis ........................................................................... 99
VIII. Conclusions and Policy Recommendations .................................................... 103
A. Emissions Standards and Model Rules for Distributed Generators.............. 104
B. New England Demand Response Initiative and Efforts to Improve Data
Collection and Coordination with ISOs.............................................................. 106
C. Current State Activities Related to the Environmental Regulation of
Distributed Generators in the Northeast ............................................................. 107
1. Connecticut ............................................................................................. 107
2. Maine ...................................................................................................... 107
3. Massachusetts ......................................................................................... 108
4. New Hampshire ...................................................................................... 108
5. New Jersey .............................................................................................. 108
6. New York................................................................................................ 109
8. Vermont .................................................................................................. 110
D. Control Technology, Cost and Other Policy Considerations........................ 110
E. Policy Recommendations for the Northeast States....................................... 112
1. Updating Emissions Standards and Permitting Requirements................ 112
2. Regulating Use of Diesel Generators in Demand Response Programs .. 114
3. Improving Regional Coordination and Data Collection ......................... 115
F. Conclusions................................................................................................... 116
Appendix A: Power Systems Research Methodology................................................ A-1
Appendix B: Power Systems Research Population Estimates for Natural Gas
Engines ........................................................................................................................... B-1
Appendix C: Power Systems Research Generator Set Estimator ............................ C-1
vii
Appendix D: Additional Background on Stationary IC Engines and Emissions ... D-1
Appendix E: ESI International Case Study Questionnaire ...................................... E-1
viii
List of Tables and Figures
Executive Summary
Table ES-1. Summary of State Permitting Requirements for Distributed Generator.......
............................................................................................................... ES-2
Table ES-2. PSR Estimates of Diesel Engines in the NESCAUM Region by Number
and Capacity.......................................................................................... ES-3
Table ES-3. Number of Permitted Engines in the Northeast States.......................... ES-4
Table ES-4. Identified Engine Populations for New York City and Fairfield County .....
............................................................................................................... ES-5
Table ES-5. Emissions Estimates for Engine Populations in New York City and
Fairfield County .................................................................................... ES-6
Table ES-6. Distributed Generator NOx, PM10 and VOC Emissions Factors.......... ES-6
Table ES-7. Estimated Emissions for 2002 NY-ISO Price Response Program in New
York City .............................................................................................. ES-6
Table ES-8. Estimated Emissions for 2002 NE-ISO Price Response Program in New
England ................................................................................................. ES-7
Table ES-9. Summary of Recommended/Adopted Distributed Generation Emissions
Standards............................................................................................. ES-10
Table ES-10. Summary of NESCAUM Recommendations ..................................... ES-11
Chapter I
Figure I-1.
Figure I-2.
Chapter II
Table II-1.
Chapter III
Table III-1.
Table III-2.
Table III-3.
Table III-4.
Table III-5.
Table III-6.
Table III-7.
Table III-8.
NOx Emissions Factors of Diesel Engines Relative to Other Types of
Generation................................................................................................... 2
PM10 Emissions Factors of Diesel Engines Relative to Other Types of
Generation................................................................................................... 3
Summary of State Permitting Requirements for Distributed Generators ... 8
PSR Estimates of Diesel Engines in the NESCAUM Region by Number
and Capacity.............................................................................................. 16
Number of Electricity-Generating Engines in NESCAUM State Permit
Records ..................................................................................................... 16
Comparison of PSR Estimated Generators and State Permit Records ..... 17
Mechanical Engines in NESCAUM State Permit Records....................... 18
PSR Estimates of Diesel Engines in Connecticut by Number and Capacity
................................................................................................................... 18
Connecticut Permit Records for Electricity Generators and Mechanical
Engines...................................................................................................... 19
Connecticut Engine Data Comparison by Number and Capacity............. 19
PSR Estimates of Diesel Engines in Maine by Number and Capacity ..... 20
ix
Table III-9.
Table III-10.
Table III-11.
Table III-12.
Table III-13.
Table III-14.
Table III-15.
Table III-16.
Table III-17.
Table III-18.
Table III-19.
Table III-20.
Table III-21.
Table III-22.
Table III-23.
Table III-24.
Table III-25.
Table III-26.
Table III-27.
Table III-28.
Maine Permit Records for Electricity Generators and Mechanical Engines
................................................................................................................... 20
Maine Engine Data Comparison by Number and Capacity...................... 21
PSR Estimates of Diesel Engines in Massachusetts by Number and
Capacity .................................................................................................... 21
Massachusetts Permit Records for Electricity Generators and Mechanical
Engines...................................................................................................... 22
Massachusetts Engine Data Comparison by Number and Capacity......... 22
PSR Estimates of Diesel Engines in New Hampshire by Number and
Capacity .................................................................................................... 23
New Hampshire Permit Records for Electricity Generators and
Mechanical Engines .................................................................................. 23
New Hampshire Engine Data Comparison by Number and Capacity...... 24
PSR Estimates of Diesel Engines in New Jersey by Number and Capacity
................................................................................................................... 24
New Jersey Permit Records for Electricity Generators and Mechanical
Engines...................................................................................................... 25
New Jersey Engine Data Comparison by Number and Capacity ............. 25
PSR Estimates of Diesel Engines in New York by Number and Capacity
................................................................................................................... 26
New York Permit Records for Electricity Generators and Mechanical
Engines...................................................................................................... 26
New York Engine Data Comparison by Number and Capacity ............... 27
PSR Estimates of Diesel Engines in Rhode Island by Number and
Capacity .................................................................................................... 27
Rhode Island Permit Records for Electricity Generators and Mechanical
Engines...................................................................................................... 28
Rhode Island Engine Data Comparison by Number and Capacity........... 28
PSR Estimates of Diesel Engines in Vermont by Number and Capacity . 29
Vermont Permit Records for Electricity Generators and Mechanical
Engines...................................................................................................... 29
Vermont Engine Data Comparison by Number and Capacity.................. 30
Chapter IV
Table IV-1.
Table IV-2.
Table IV-3.
New York City PSR Survey Results by Engine Fuel and Application .... 33
New York City PSR Survey Results by Engine Size and Capacity ......... 33
New York City PSR Survey Generation Totals in MWh/yr by Fuel and
Application................................................................................................ 33
Table IV-4. Fairfield County PSR Survey Results by Engine Fuel and Application... 34
Table IV-5. Fairfield County PSR Survey Results by Engine Size and Capacity........ 34
Table IV-6. Fairfield County PSR Survey Generation Totals in MWh/yr by Fuel and
Application................................................................................................ 35
Table IV-7. New York City DEP Permitted Emergency Engines................................ 36
Table IV-8. Total Engine Inventory for New York City.............................................. 36
Table IV-9. Connecticut DEP Permitted Engine Summary for Fairfield County ........ 37
Table IV-10. Total Engine Inventory for Fairfield County ............................................ 37
x
Chapter V
Table V-1.
Table V-2.
Table V-3.
Table V-4.
Table V-5.
Table V-6.
Table V-7.
Table V-8.
Table V-9.
Table V-10.
Figure V-1.
Distributed Generator NOx, PM10 and VOC Emissions Factors.............. 43
Estimated New York City Emissions from Surveyed and Permitted
Engines According to PSR Survey Results .............................................. 44
Estimated Fairfield County Emissions from Surveyed and Permitted
Engines According to PSR Survey Results .............................................. 45
Emissions Estimates for 2002 New York ISO Price-Responsive Load
Program in New York City ....................................................................... 47
Emissions Estimates for 2001 New England ISO Demand Response
Programs ................................................................................................... 48
New England 2002 Program Enrollment.................................................. 49
Emissions Estimates for 2002 New England ISO Price Response Program
................................................................................................................... 50
Distribution of 2002 NE ISO Price Response Events and Emissions, by
State........................................................................................................... 50
Generation and Estimated Emissions for 2002 NE ISO Price Response
Program in Southwest CT......................................................................... 50
New England Energy Clearing Price (ECP) New York City Location
Based Marginal Price (LBMP) Data......................................................... 53
Hours that Electricity Prices Exceeded 8¢ and 10¢/kWh in 2001 and 2002
................................................................................................................... 54
Chapter VI
Figure VI-1. Permitted Stationary Diesel Engine Database Manufacturer Listing ....... 56
Table VI-1. Top Twelve States with Combined Heat & Power Sites Using Stationary
IC Engines................................................................................................. 58
Figure VI-2. An Example of Multiple Engine Usage (Tacoma Power, CAT 3516B
Diesel Engines with SCR) ....................................................................... 59
Table VI-2. Typical Emissions Factors for Stationary IC Engines .............................. 60
Figure VI-3. Control Capabilities of Catalyst Control Technology vs. A/F Ratio ........ 61
Figure VI-4. Mechanical Filtration of a Wallflow DPF ................................................ 64
Table VI-3. Specifications of a 13-Mode Engine Test Cycle....................................... 66
Figure VI-5. 13-Mode Steady-State Test Cycle for DPF PM Emissions Reduction..... 66
Table VI-4. Reductions in PAH Emissions for Two DPF Systems ............................. 67
Figure VI-6. Diesel Oxidation Catalysts........................................................................ 68
Figure VI-7. DOC Performance..................................................................................... 69
Figure VI-8. 13-Mode Steady-State Test Cycle for DOC PM Emissions Reductions .. 70
Table VI-5. Reductions in PAH Emissions for DOCs ................................................. 71
Figure VI-9. Schematic of Crankcase Emission Control............................................... 72
Figure VI-10. Selective Catalytic Reduction ................................................................... 73
Figure VI-11. 13-Mode Steady-State Test Cycle for SCR NOx Emissions Reductions . 73
Figure VI-12. Control Performance of a Combined DPF/SCR System .......................... 74
Figure VI-13. Exhaust Gas Recirculation........................................................................ 75
xi
Chapter VII
Table VII-1.
Table VII-2.
Table VII-3.
Table VII-4.
Table VII-5.
Table VII-6.
Table VII-7.
Table VII-8.
Table VII-9.
Table VII-10.
Table VII-11.
Table VII-12.
Table VII-13.
Table VII-14.
Table VII-15.
Table VII-16.
Table VII-17.
Table VII-18.
Table VII-19.
Table VII-20.
Figure VII-1.
Figure VII-2.
Summary of Case Study Facilities............................................................ 78
Case Study 1: Engine and Emission Control Summary............................ 81
Case Study 1: Emissions Reduction Summary......................................... 81
Case Study 1: Annual DPF Costs and Emissions ..................................... 83
Case Study 2: Engine and Emission Control Summary............................ 84
Case Study 2: Emissions Reduction Summary......................................... 85
Case Study 2: Annual DPF Costs and Emissions ..................................... 86
Case Study 2: Annual SCR Costs and Emissions..................................... 87
Case Study 3: Engine and Emission Control Summary............................ 88
Case Study 3: Emissions Reduction Summary......................................... 88
Case Study 3: Annual DPF Costs and Emissions ..................................... 89
Case Study 4: Engine and Emission Control Summary............................ 90
Case Study 4: Emissions Reduction Summary......................................... 91
Case Study 4: Annual DOC Costs and Emissions .................................... 92
Case Study 5: Engine and Emission Control Summary............................ 93
Case Study 5: Emissions Reduction Summary......................................... 94
Case Study 5: Annual DPF Costs and Emissions ..................................... 95
Case Study 6: Engine and Emission Control Summary............................ 96
Case Study 6: Emissions Reduction Summary......................................... 97
Case Study 6: Annual DPF Costs and Emissions ..................................... 98
Cost-Effectiveness Versus Annual Operating Hours.............................. 100
Control Cost-Effectiveness for Actual Operation at Kings County and
NASSCO................................................................................................. 101
Chapter VIII
Table VIII-1. Summary of Recommended/Adopted Distributed Generation Emissions
Standards................................................................................................. 105
Table VIII-2. Summary of NESCAUM Recommendations ......................................... 117
xii
List of Abbreviations and Acronyms
lb/MMBtu – pounds per million Btu
lb/MWh – pounds per megawatt-hour
LBMP – Location Based Marginal Price
MACT – Maximum Achievable Control
Technology
MMBtu – million Btu
MW – megawatt (= 1,000 kW)
MWh – megawatt-hour
NAICS – North American Industry
Classification System
NEDRI – New England Demand Response
Initiative
NESCAUM – Northeast States for
Coordinated Air Use Management
N2 - nitrogen
NMHC – non-methane hydrocarbons
NOx – nitrogen oxides
NRDC – Natural Resources Defense
Council
NSR – New Source Review
O2 – oxygen
OAQPS – Office of Air Quality Planning
and Standards (EPA)
OTC – Ozone Transport Commission
PAH – polycyclic aromatic hydrocarbons
PM – particulate matter
PM2.5 – PM < 2.5 microns in diameter
PM10 – PM < 10 microns in diameter
ppm – parts per million
ppmvd – parts per million volume, dry
PJM – ISO for the Mid-Atlantic states
PRP – price response program
PSR – Power Systems Research
PTE – potential to emit
RACT – Reasonably Available Control
Technology
RAP – Regulatory Assistance Project
RBL – RACT/BACT/LAER
SCR – selective catalytic reduction
SI – spark ignition
SIC – Standard Industrial Classification
SMAQMD – Sacramento Metropolitan Air
Quality Management District
SOTA – state of the art
SOX/SO2 – sulfur oxides/sulfur dioxide
TPY – tons per year
TWC – three-way catalyst
VOC – volatile organic compounds
ºC/ºF – degree Celsius/Fahrenheit
A/F – air to fuel ratio
BACT – Best Available Control Technology
bhp – brake horsepower
bhp-hr – brake horsepower-hour
Btu – British thermal unit
CARB – California Air Resources Board
CHP – combined heat and power
CI – compression ignition
CO – carbon monoxide
CRF – capital recovery factor
DEC – Department of Environmental
Conservation
DEM – Department of Environmental
Management
DEP – Department of Environmental
Protection
DES – Department of Environmental
Services
DG – distributed generation
DOC – diesel oxidation catalyst
DOE – Department of Energy
DPF – diesel particulate filter
DR – demand response
ECP – energy clearing price
EGR – exhaust gas recirculation
EDRP – emergency demand response
program
EPA – U.S. Environmental Protection
Agency
FERC – Federal Energy Regulatory
Commission
FTP – Federal Test Procedure
g – gram
GPDG – General Permit for Distributed
Generation
GPEE – General Permit for Emergency
Engines
GW – gigawatt (=1,000 MW)
HAP – hazardous air pollutant
HC – hydrocarbons
hp – horsepower
IC – internal combustion
ISO – Independent System Operator
kW – kilowatt
kWh – kilowatt-hour
LAER – Lowest Achievable Emission Rate
lb – pound
xiii
Executive Summary
Stationary diesel internal combustion (IC) engines constitute a significant component of
the nation’s electricity generating infrastructure. Estimates of installed diesel generator
capacity in the United States range as high as 350,000 units totaling more than 127
gigawatts (GW);1 estimates developed for this report suggest that the total population of
diesel generators in the Northeast could include well over 30,000 units with a combined
capacity exceeding 10 GW. Historically, the vast majority of these engines has been used
primarily or exclusively to provide back-up power in emergency (i.e. outage) situations
and in some cases to reduce reliance on grid-supplied electricity during periods of peak
demand. Consequently, most diesel generators have been operated infrequently and have
not been subject to the kinds of environmental regulation applicable to large centralstation power plants.
More recently, emerging concerns about system reliability and price volatility in
deregulated electricity markets have prompted interest in making greater use of all forms
of distributed generation capacity to lower demand for grid-supplied electricity during
high price and peak use periods. While this interest may eventually lead to increased
reliance on fuel cells, microturbines, renewable power and other advanced distributed
generation technologies, diesel IC engines are likely to remain by far the most ubiquitous
distributed generating resource available in the short term. Therefore, any increase in the
near-term use of these resources in general, must raise environmental concerns related to
the operation of diesel engines in particular. Most diesel IC engines emit high levels of
pollutants such as nitrogen oxides (NOx), a key ingredient in the formation of groundlevel ozone, and particulate matter (PM). In addition, diesel exhaust contains numerous
toxic and potentially carcinogenic components. In fact, emissions rates per unit of
electrical output for diesel IC engines are typically several times higher than those of
conventional fossil fuel power plants and orders of magnitude higher than those of the
cleanest conventional central-station generating technologies, such as large combinedcycle natural gas turbines.
State and federal regulators recognize that existing environmental policies will need to be
updated or augmented to ensure that a new generation of cleaner distributed technologies
becomes available in the future and to manage any adverse impacts from the existing
generator population in the transition, especially if market conditions and/or government
policies prompt increased use of this capacity in the near-term. Unfortunately, the
situation is complicated by a shortage of reliable information on the current population of
small distributed generators. The chief purpose of this study was therefore to begin
developing a more complete inventory of the numbers and types of diesel IC engines that
exist in the eight-state NESCAUM region.2 In addition, the study reviews current state
policies concerning the permitting and operation of diesel generators, provides
1
National diesel generator population estimate from Natural Resources Defense Council (NRDC) report
Distributed Resources and Their Emissions: Modeling the Impacts, Greene, Hammerschlag, and Keith,
2001. These figures were provided to NRDC by Power Systems Research (see Footnote 4).
2
Specifically, the states of Connecticut, Maine, Massachusetts, New Hampshire, New Jersey, New York,
Rhode Island and Vermont. All of these states are members of NESCAUM.
ES-1
preliminary estimates of emissions impacts associated with current levels of diesel
generator operation, reviews control technology options (including case studies of several
actual installations) and provides a number of specific policy recommendations.
A. Summary of Current Permitting Requirements for Diesel IC
Engines in the Northeast States
Distributed generators, and stationary internal combustion engines in general, are for the
most part regulated and permitted at the state and local level. Table ES-1 summarizes the
different permitting requirements applicable to electricity generating engines in the eight
NESCAUM states. As seen in Table ES-1, most states make a distinction between
emergency and non-emergency engines. Emergency engines are often exempt from
emissions limits or control technology requirements, however their operation is usually
Table ES-1
Summary of State Permitting Requirements for Distributed Generators
State
Non-Emergency Enginesa
Threshold
Emergency Engines
b
Requirements
Threshold
Restrictions
Demand Responsed
500 hrs/yr and
no PRP; EDRP in SW
CT: permit-by-rule
maximum of 5 TPY
CT for add'l 300 hrs/yr,
SW CT: 50 hp (37
NOx, 5 TPY CO, 3 TPY
kW)
only nat. gas or ULSDe
PM, and 3 TPY SO2
CT
PTE 15 TPYc of any
criteria pollutant
ME
5 MMBtu/hr (approx.
SCR over 20 TPY NOx, 0.5 MMBtu/hr
500 kW), 0.5 MMBtu/hr BACT case-by-case, on- (approximately 50
if at major source
road diesel
kW)
500 hrs/yr
no additional
restrictions
MA
3 MMBtu/hr (approx.
300 kW), smaller if at
facility with other
permitted engines
3 MMBtu/hr permitby-rule, over 10
MMBtu/hr case-bycase BACT
300 hrs/yr, cannot
create a "condition of air
pollution," must have a
noise muffler
no PRP; may run once
ISO has called for
voltage reductions (OP4 step 12 or 14)
no threshold
500 hrs/yr, limit sulfur
content of diesel, and
limit emissions
neither type of DR for
emergency engines
NH
NJ
BACT, LAER based on
emissions
case-by-case BACT
1.5 MMBtu/hr (150 kW)
diesel, 10 MMBtu/hr (1 over 400 kW may
MW) natural gas, PTE require RACT
25 TPYc NOx
BACT for new/modified;
existing diesel engines
1 MMBtu/hr
require 8g/bhp-hr NOx
(approximately 100 kW) (being revised to 2.3
g/bhp-hr)
1 MMBtu/hr
no control if PTE NOx is neither type of DR for
(approximately 100
less than 25 TPY
emergency engines
kW)
NY
NY: no threshold, NY: 500 hrs/yr, no
NY: 300 kW, 160 kW if diesel engines are not
non-att., NYC: 280 kW, allowed to participate in NYC: over 280 kW permits, NYC: register
must register
but no restrictions
33 kW if diesel
PRPd
RI
500 kW diesel,
1 MW natural gas
VT
450 hp (337 kW)
BACT based on
emissions
must meet EPA's nonroad standards
no threshold
500 hrs/yr, 0.3% sulfur
diesel fuel
no threshold
200 hrs/yr
no PRP; EDRP less
than 200 hrs/yr, 30 ppm
sulfur diesel fuel
required
neither type of DR for
emergency engines
neither type of DR for
emergency engines
a
non-emergency engines are not restricted from participating in demand response programs, except as noted in NY
abbreviations: BACT=Best Available Control Technology; MACT=Maximum Achievable Control Technology; RACT=
Reasonably Available Control Technology; LAER=Lowest Achievable Emission Rate; SCR=Selective Catalytic Reduction;
SOTA=State of the Art
b
c
PTE=potential to emit, and TPY=tons per year
demand response (DR) programs include the emergency demand response program (EDRP) which is called by the ISO in the
event of an imminent capacity shortfall, and the price response program (PRP) in which customers respond to high prices
d
e
ultra-low sulfur diesel fuel
ES-2
strictly limited to certain situations and a maximum number of hours (typically from 200
to 500 hours) per year. By contrast, non-emergency engines are typically regulated down
to smaller sizes and to more stringent emissions control requirements. Table ES-1 also
notes different state policies regarding the eligibility of emergency generators to
participate in formal “demand response” programs.3 In some states, emergency units are
allowed to operate under emergency demand response programs – which are invoked at
times of imminent supply shortfalls to avert loss of grid power – while in others,
operation of emergency units remains constrained to actual outage situations only.
Emergency units are generally precluded in all Northeast states from participating in
price response programs which are designed to reduce system load during periods of high
prices (as opposed to supply shortages).
B. Northeast States Distributed Generator Inventory
In an attempt to develop more precise inventories of the population of diesel and other
distributed generators in the Northeast, NESCAUM relied on estimates developed by a
consultant, Power Systems Research4 (PSR), together with information gathered from
individual state permit records. The PSR estimates were derived using a methodology
developed from national sales data and field surveys that correlates engine population to
the numbers and types of businesses present in a given geographic area. Table ES-2
summarizes the PSR population estimates for the NESCAUM region. Engine totals are
shown by number and by capacity, sorted by size range and type of application.
Table ES-2
PSR Estimates of Diesel Engines in the NESCAUM Region by Number and Capacity
Number Totals Emergency Peak Baseload Total
1,768
0
0
25-50 kW
1,768
50-100 kW
5,798
1,375
107
7,280
100-250 kW
9,226
2,236
95
11,557
250-500 kW
5,918
1,231
7
7,156
500-750 kW
1,296
316
47
1,659
750-1000 kW
1,164
292
51
1,507
641
677
39
1000-1500 kW
1,357
1500+ kW
1,073
284
37
1,394
Total
26,884
6,411
383
33,678
Capacity Totals (MW) Emergency Peak Baseload Total
59
0
0
25-50 kW
59
50-100 kW
462
114
9
584
100-250 kW
1,564
371
14
1,949
250-500 kW
2,126
443
3
2,572
500-750 kW
801
196
29
1,026
750-1000 kW
921
230
40
1,191
769
837
48
1000-1500 kW
1,654
1500+ kW
2,053
615
68
2,736
Total
8,756
2,805
211
11,772
Overall, PSR estimates that a total of 33,678 diesel IC engines, with a total capacity of
11,772 MW, are currently installed in the eight NESCAUM states. Of this population, a
very large fraction, estimated at 80% of the engines (accounting for 74% of total MW
capacity), is designated for emergency use.
Information on the diesel generator population from state permitting records is
summarized in Table ES-3, below. Because the amount of information and unit-level
3
With the exception of New York (as detailed in Table ES-1), non-emergency engines are not restricted
from participation in demand response programs.
4
Power Systems Research (PSR) is a market research company for the engine industry.
ES-3
detail available from each state varied, the size or capacity distribution for some engines
needed to be estimated for at least part of the permitted population in some states.
Table ES-3
Number of Permitted Engines in the Northeast States
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
CT
112
208
411
321
273
144
153
99
1,721
ME
2
78
184
158
64
28
36
28
578
MA
11
13
278
156
138
73
160
275
1,104
NH
1
2
65
126
71
39
47
9
360
NJ
4
120
1,432
1,247
927
837
698
558
5,823
NY
26
93
337
410
272
201
175
148
1,662
RI
0
0
4
1
20
11
11
25
72
VT
0
9
18
17
7
2
10
3
66
TOTAL
156
523
2,729
2,436
1,772
1,335
1,290
1,145
11,386
Comparing Tables ES-2 and ES-3 suggests that current state permit records capture
approximately one-third of the total number of diesel IC generators estimated by PSR for
the NESCAUM region. As one would expect, based on the fact that many smaller
engines fall below current state permitting thresholds, the agreement between PSR’s
estimates and state permitting records generally improves in the larger engine size
categories (i.e. > 250 kW). However, it is important to note that PSR’s estimation
methodology is inherently inaccurate. A closer comparison to state permit records (as
described in detail in Chapter III of this report) reveals both that the discrepancies are
significant and that there is no particular pattern to the variance between states’ permit
data and PSR’s population estimates for different engine size categories. This suggests
either that the PSR estimates are inexact, or that available state permit records are
incomplete, or some combination of both.
C. Results of Detailed Engine Surveys in New York City and Fairfield
County, Connecticut
As a follow-up and complement to the engine inventory efforts described above, PSR
conducted detailed telephone surveys to obtain information on the distributed engine
population in New York City and Fairfield County, Connecticut. These areas were
selected because both face severe transmission constraints and have been the focus of
recent efforts by electric system operators to encourage customer-side demand responses
during periods of peak electrical demand. The results obtained by PSR in the telephone
surveys were compared to available information from state or city permit records for
these two areas. This analysis differed from the statewide estimates above in that it was
possible to determine from the telephone survey data specifically which engines were (or
were not) included in state permit files.
In general, the comparison revealed surprisingly little overlap between the telephone
survey results and available permit records, even for many of the larger engines identified
ES-4
by PSR. The percentage of surveyed engines for which there was no record in state or
city databases was over 50% in the case of New York City and over 80% in the case of
Fairfield County. Clearly, a significant portion of this discrepancy can be explained by a
lack of data for those engines for which no permit is required (or available) because they
are operating under an emergency exemption or permit-by-rule exclusion, or simply fall
below applicable permitting thresholds. Nevertheless, this result suggests that neither the
survey/estimation methodology used by PSR, nor current state permitting systems can be
relied upon for comprehensive coverage of the existing population of small generators
Table ES-4 summarizes the combined population of diesel generators identified through
either the survey and/or available permit records for the two survey areas.
Table ES-4
Identified Engine Populations for New York City and Fairfield County
Engine Size
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
New York City
Fairfield County, CT
Total
% of
Capacity
Total
% of
Capacity
Number Engines Total (MW) Number Engines Total (MW)
26
1.2%
1
29
5.2%
1
123
5.5%
8
108
19.3%
7
426
19.1%
70
135
24.1%
21
509
22.8%
178
148
26.4%
48
389
17.4%
229
66
11.8%
37
328
14.7%
277
19
3.4%
16
319
14.3%
346
29
5.2%
32
116
5.2%
211
26
4.6%
74
2,236
100%
1,320
560
100%
235
D. Emissions Estimates
Because small diesel generators are often located near or in densely populated urban
areas and because their emissions tend to be released closer to the ground, operation of
these engines – especially during peak demand hours (which typically occur on hot
summer days when air quality is already poor) – poses particular public health concerns.
Because no mechanism has existed to comprehensively track these units, empirical data
on their historic emissions have been scarce. The telephone survey portion of this study
enabled NESCAUM to collect information on the actual operation of engines known to
exist in New York City and Fairfield County, Connecticut. Table ES-5 summarizes our
estimate of annual emissions associated with the total engine population identified
through either the telephone surveys or permitting records in these two areas. Emissions
totals are calculated using emissions factors and average operating hours for different
types and sizes of engines as reported by owners contacted in the telephone surveys. The
emissions factors used for purposes of this analysis are summarized in Table ES-6.5
5
The emissions factors shown in Table ES-6 are from the Sacramento Metropolitan Air Quality
Management District’s (SMAQMD) September 2002 Internal Combustion Engine Manual. The SMAQMD
Manual uses AP-42 emissions factors, except where emissions tests conducted by California for purposes
of Best Available Control Technology (BACT) determinations revealed higher emissions factors (see
further discussion in Chapter V).
ES-5
Table ES-5
Emissions Estimates for Engine Populations in New York City and Fairfield County
Number of
Engines
Diesel
1,652
Natural Gas
549
Gasoline
35
Total
2,236
New York City
Fairfield County, CT
PM10
PM10
VOC Number of
VOC
NOx
NOx
MWh/yr
MWh/yr
(tons/yr) (tons/yr) (tons/yr) Engines
(tons/yr) (tons/yr) (tons/yr)
379,620 6,480
260
580
447
313,040 5,220
190
470
144,720 2,320
30
90
90
3,700
60
0.8
2.5
7,360
50
5
110
23
2,820
20
1.4
40
531,700 8,850
295
780
560
319,560 5,300
192
513
Table ES-6
Distributed Generator NOx, PM10 and VOC Emissions Factors
Engine Type
Diesel < 600 hp
Diesel > 600 hp
Natural Gasa
Gasoline
a
b
NOx
g/hp-hr
lb/MWh
14.06
41.47
10.86
32.04
b
b
10.89
32.12
5.00
14.72
PM10
g/hp-hr
lb/MWh
1.00
2.95
0.32
0.94
b
b
0.15
0.45
0.33
0.97
VOC
g/hp-hr
lb/MWh
1.14
3.36
b
b
1.00
2.95
b
b
0.43
1.27
9.79
28.88
Emissions factors represent an average for three types of natural gas engines. Additional detail in Table VI-2.
SMAQMD emission factors used in place of AP-42, see Footnote 5.
Note that emissions were not calculated for the broader engine population estimates
developed by PSR for the NESCAUM region as a whole. The results of such a
calculation would be questionable due to the higher level of uncertainties involved.
In addition, NESCAUM analyzed the emissions impacts associated with engine operation
under formal demand response programs sponsored by the New York and New England
grid operators or independent system operators (ISO’s) in recent years. Tables ES-7 and
ES-8 summarize the results of this analysis for the summer 2002 economic or “price
response” programs,6 using the emissions factors shown in Table ES-6.
Table ES-7
Estimated Emissions for 2002 NY-ISO Price Response Program in New York City
Date
04/17/02
04/18/02
07/30/02
08/14/02
Total
6
MWh Generated
154.6
191.5
267.7
251.3
865.1
NOx (tons)
2.48
3.07
4.29
4.03
13.86
PM10 (tons)
0.07
0.09
0.13
0.12
0.41
Neither ISO invoked its emergency demand response program in 2002.
ES-6
VOC (tons)
0.23
0.28
0.39
0.37
1.28
Table ES-8
Estimated Emissions for 2002 NE-ISO Price Response Program in New England
State
Connecticut
Massachusetts
Maine
New Hampshire
Rhode Island
Vermont
Total
MWh
Curtailed
272.25
88.38
62.55
7.48
3.44
0.04
434.13
MWh
Generated
421.97
27.07
81.82
0.00
0.00
72.53
603.39
NOx
(tons)
6.76
0.43
1.31
0.00
0.00
1.16
9.67
PM10
(tons)
0.20
0.01
0.04
0.00
0.00
0.03
0.28
VOC
(tons)
0.62
0.04
0.12
0.00
0.00
0.11
0.89
In general, this preliminary analysis suggests that additional emissions impacts associated
with the use of stationary diesel generators in recently introduced formal demand
response programs have to date been small – on an annual basis – relative to state and
local inventories of emissions from all pollutant sources. Moreover, emissions from the
existing operation of larger, non-emergency engines for peak-shaving and baseload
purposes are likely to dwarf any near-term increase in emissions associated with the use
of diesel generators under the formal demand response programs being introduced or
augmented by grid operators. In New Hampshire, for example, no new generation
occurred under the New England ISO’s formal price response program in 2002.
However, the New Hampshire Department of Environmental Services had documented a
significant increase in NOx emissions from stationary IC engines in the 1990s,
presumably as the result of the increased operation of non-emergency engines in response
to other market factors.7
In any case, the potential for increased reliance on distributed generation resources in
general, and smaller diesel-powered IC engines in particular, could grow in the future if
demand response programs are substantially expanded and/or if increasing numbers of
large customers with on-site generating capacity opt to purchase electricity on the spot
market or are otherwise exposed to real-time spot market prices. In that case, the
economic incentives for switching from grid-supplied electricity to on-site generation
might be substantial during some hours of the year. Our preliminary analysis indicates
that wholesale electricity prices in the New England and New York power pools have
risen above 8 cents per kilowatt-hour (the level above which it might be economic to
switch to on-site generation) on the order of 100 to 200 hours per year in recent years.
This range may increase in future years as demand catches up to the substantial new
central-station generating capacity that was added in the region between 1999 and 2002.
Meanwhile high prices are already occurring much more often (on the order of 500 hours
per year) in transmission-constrained load pockets such as New York City.
7
Specifically, NOx emissions from fossil-fuel fired IC engines grew from 1.4% to 14% of New
Hampshire’s total ozone season NOx inventory from electric generating units between 1993 and 1999. In
just the 3 years between 1996 and 1999, estimated NOx emissions from these engines more than doubled –
from 243 tons in the 1996 ozone season to 576 tons in the 1999 ozone season. Moreover, the NH
Department of Environmental Services was aware of at least two instances where multiple IC engines were
installed to reduce electricity costs. These developments prompted NH to adopt new emissions rules for IC
engines in 2001.
ES-7
It is important to emphasize that the more important emissions and public health concerns
related to diesel generators probably have less to do with their NOx emissions or their
contribution to the overall emissions inventories typically used for traditional attainment
planning purposes. Rather, short-term, highly localized impacts associated with
particulate matter and toxic emissions are likely to constitute the most significant air
quality and health concerns relevant to the use of these engines.
E. Emissions Control Options for Stationary Diesel Generators
As described in Chapter VI, various control technologies exist that can substantially
reduce diesel engine emissions. For example, particulate emission control options include
filters – which can provide 80-90% levels of control – and oxidation catalysts. The latter
technology achieves more modest PM reductions (about 20%), but is also less costly. In
addition, both filter and catalyst controls can provide significant (80-90%) reductions in
hydrocarbon (HC) emissions, including toxic hydrocarbons, as well as carbon monoxide
(CO) reductions. Selective catalytic reduction (SCR) technology has been successfully
applied to large diesel engines where it can achieve NOx reductions of 80-90% or more.
However, these systems are more cost-effective for large engines that operate frequently.
Less costly NOx control strategies include injection timing adjustments; these generally
provide more modest reductions (on the order of 10-20%).
Chapter VII reviews six case studies which provide useful information on the technical
feasibility and cost effectiveness of various diesel engine emission control options. Four
of the case studies examine the use of particulate filters, one analyzes the combined
application of particulate filters and SCR, and one describes an installation that employed
oxidation catalysts. Based on the case studies, the capital cost of a filter system ranges
from about $45,000 to about $120,000. The sole diesel oxidation catalyst installation, by
contrast, had a more modest capital cost of $25,000. The capital cost of the SCR
installation was estimated to be just under $180,000. It should be cautioned, however,
that all the case study installations involved generally large engines (in the range of
1,000-3,000 hp, approximately equivalent to 750-2,250 kW). Hence, the applicability of
these cost estimates to smaller engines may be somewhat uncertain.
The chief issue for emissions control of diesel IC generators, particularly in the case of
smaller engines, is one of cost and cost-effectiveness, rather than technical feasibility. If
an engine runs for only a few tens of hours in a year, a given control technology will
likely remove only relatively small amounts of emissions (say, compared to a centralstation power plant) for a given capital cost. Not surprisingly, data from the case studies
indicate that cost effectiveness – as measured by the conventional metric of tons removed
per dollar of control cost – improves as operating hours increase. At 500 hours per year
of operation, control costs for the filter and catalyst installations in the case studies range
from $2,000-$90,000 per combined ton of PM, CO and HC reductions; at 2,000 hours per
year, the costs fall to $1,000-$23,000 per ton. For the SCR case study, estimated control
costs were just under $8,000 per ton of NOx removed. Cost-effectiveness also varies with
the size of the engine involved.
ES-8
In this context, innovative regulatory approaches can provide attractive alternatives to
mandating costly retrofit controls. For example, New Hampshire has introduced a
program of emissions fees for small diesel generators. Fees are imposed per ton of NOx
generated and are scheduled to increase over time. Similar incentive programs could be
used to promote less polluting engines elsewhere, with resulting revenues applied to
research, development and demonstration and to support available cleaner technologies.
Another option would be to require distributed generators to obtain pollution allowances,
especially where emissions budget and trading programs already exist.
Finally, new federal standards recently introduced for non-road diesel engines used in
farming, construction and industrial activities are relevant to the future regulation of
stationary engines. The standards require substantial reductions in NOx and PM
emissions; in addition, they establish fuel content requirements (i.e. sulfur limits) in
recognition of the fact that many advanced control technologies require low-sulfur fuel.
Importantly, low-sulfur fuel will provide immediate emissions reductions in the existing
engine fleet, in addition to any reductions that are gradually achieved by the introduction
of new engine models. Meanwhile, efforts are also underway to develop and apply
retrofit emissions control technologies for existing on-road and non-road engines, many
of which are likely to be similarly applicable to existing stationary diesel generators.
F. Current State, Regional and National Policy Initiatives Related to
Diesel IC Generators
A number of policy initiatives aimed at regulating diesel generators specifically and
promoting cleaner distributed generation alternatives more broadly have recently been
undertaken at the state, regional, and national levels. At the state level, several states –
notably Texas and California – have recently adopted more stringent emissions standards
for stationary diesel generators. Similarly, several Northeast states are currently in the
process of reviewing and updating their standards and/or permitting requirements; while
others plan to do so in the near future. At the regional and national levels, the Ozone
Transport Commission (OTC) and the Regulatory Assistance Project (RAP) have
recently developed model rules for the regulation of diesel and other distributed
generators. The OTC model rule recommends fuel-specific NOx standards for all nonemergency natural gas and diesel engines and suggests a number of operational and
record-keeping requirements or restrictions for emergency engines. By contrast, the RAP
model rule is aimed at new engines only and recommends the phased introduction of
progressively more stringent output-based, fuel-independent emissions standards for
several pollutants. Table ES-9 summarizes the specific recommendations in the RAP and
OTC model rule, as well as new requirements adopted in Texas and California.
ES-9
Table ES-9
Summary of Recommended/Adopted Distributed Generation Emissions Standards
NOx (Ozone
Attainment Areas)
Regulation
Emission Limits (lb/MWh)
NOx (Ozone NonVOCs
CO
Attainment Areas)
PM
CO2
a
OTC Model Rule - new and in use engines (emissions factors converted from g/bhp-hr)
Natural Gas (except emergency)
4.4
4.4
Diesel (except emergency)
6.8
6.8
RAP Model Rule - new engines
After January 1, 2004
4
0.6
10
0.7
1900
After January 1, 2008
1.5
0.3
2
0.07
1900
After January 1, 2012
0.15
0.15
1
0.03
1650
California Air Resources Board
Distributed Generation Certification Rule for New Non-Emergency Engines
After January 1, 2003
0.5
0.5
6
1
fuel req.
After January 1, 2007
0.07
0.07
0.1
0.02
fuel req.
b
b
Airborne Toxics Control Measure for New and In-Use Stationary IC Engines (DRAFT 6/5/03)
New diesel engines > 50 hp
(converted from g/bhp-hr)
c
Baseload power
off-road standards apply
0.03
Emergency power
Exisiting diesel engines > 50 hp
off-road standards apply
0.44
d
c
Baseload power
NOx, CO and VOCs not to increase > 10% to meet PM limits
Emergency power
NOx, CO and VOCs not to increase > 10% to meet PM limits
e
e
f
0.03
d
1.48
Texas - new engines
Before January 1, 2005
less than 300 hrs/yr
21
1.65
more than 300 hrs/yr
3.11
0.47
less than 300 hrs/yr
21
0.47
more than 300 hrs/yr
3.11
0.14
After January 1, 2005
a
The OTC Model Rule offers three compliance options: 1) meet the NOx emission limit specified above, 2) meet a
percentage reduction of NOx emissions specific to the type of engine, and 3) purchase of NOx allowances.
b
PM emission limit corresponding to natural gas with fuel sulfur content of no more than 1 grain/100 standard cubic foot.
c
Engine must meet model year off-road compression-ignition engine standards, or Tier 1 off-road certification standards.
d
Allowable hours of operation increase as the emissions factor of an engine decreases.
e
Many engines may require control technology to meet the PM limits set in this rule, and these technologies must not
increase the emisisons of NOx, CO or VOC by more than 10%.
f
Engines can meet this standard or reduce PM emissions by 85%.
Another recent regional initiative that is addressing issues relating to diesel generators is
the New England Demand Response Initiative (NEDRI), a multi-stakeholder process
aimed at developing policy recommendations to promote demand response more broadly
in New England. State air agency officials have been participating actively in NEDRI,
which has already endorsed a number of specific policies aimed at ensuring that on-site
generators participating in future New England demand response programs have
appropriate permits and provide information to the ISO and state regulators. These new
information collection and record-keeping practices should help address current data gaps
and provide state and federal authorities with a better foundation for estimating emissions
impacts of demand response programs and developing future policies related to the
regulation of distributed generators.
ES-10
G. Policy Recommendations
This report closes with a number of policy recommendations for states to consider
concerning the future use and regulation of diesel IC engines used to generate electricity.
The recommendations are divided into three categories: (1) updating emissions standards
and air permitting requirements; (2) regulating use of diesel generators in demand
response programs; and (3) improving regional coordination and data collection. Specific
policy recommendations are summarized in Table ES-10, below.
Table ES-10
Summary of NESCAUM Recommendations
Updating Emissions
Standards & Permitting
Requirements
•
•
•
•
Regulating Use of Diesel
Generators in Demand
Response Programs
•
•
•
•
•
Improving Regional
Coordination and Data
Collection
•
•
•
Review adequacy of current permitting size thresholds and
requirements, especially in light of new health concerns associated
with localized exposure to diesel exhaust.
Update requirements for existing peak, baseload and emergency
generators accordingly, especially for those units eligible to participate
in ISO or utility-sponsored demand response programs.
Consider additional fuel requirements (e.g. use of low or ultra-low
sulfur fuel) for diesel generators, especially for operation in demand
response or other non-emergency applications.
Adopt stringent output-based emissions standards for new distributed
generators, which – in the case of combined heat and power systems –
appropriately account for useful thermal, as well as electrical output.
Limit participation in non-emergency economic (price-driven) demand
response programs to generators that meet minimum emissions control
requirements (e.g. BACT-level controls, RAP model rule, etc.)
Limit participation of emergency generators to emergency demand
response programs, subject to additional requirements as deemed
appropriate under recommendations above.
Clarify regionally consistent definition of “emergency”.
Consider appropriateness of restrictions and/or additional emissions
control or operating requirements (e.g. fuel sulfur requirements) for
diesel generators participating in demand response programs.
Continue implementing NEDRI recommendations concerning the
obligation of demand response participants to verify permit status and
provide unit specific information, together with the ISO’s obligation to
provide detailed information about program outcomes on a regular
basis. The information provided must be specific enough to allow
regulators to make a reliable assessment of associated environmental
and public health impacts.
Promote more regional consistency in permitting requirements and
emissions standards for new and existing distributed generators.
Promote more regional consistency in state record-keeping practices,
with the aim of eventually developing integrated, user-friendly
information databases.
Continue to develop and refine inventories of generator population and
potential emissions impacts. Promote regional consistency in related
policies (e.g. interconnection standards, pricing policies, other air
regulatory programs, etc.)
ES-11
I.
Introduction
Stationary diesel internal combustion engines constitute a largely unseen, but significant
component of the nation’s electricity generating infrastructure. Estimates of installed
diesel generator capacity in the United States range as high as 350,000 units totaling
more than 127 gigawatts (GW).1 That compares to more than 5,000 central station power
plants with a combined generating capacity of more than 850 GW.2 Although the
combined generating capacity of these stationary diesel engines is significant, the vast
majority – especially of the small engines – are used primarily or exclusively to provide
back-up power in emergency (i.e. outage) situations and in some cases to reduce
consumption of grid-supplied electricity during periods of peak demand. Consequently,
most small diesel generators run relatively infrequently and provide only limited amounts
of power. As a result, they have generally not been subject to the kinds of environmental
regulation applicable to large central station power plants.
Despite their small size and typically infrequent operation, small diesel generators and
other distributed resources play an important role in the nation’s electric power system –
from a reliability standpoint and because they provide some customers with the ability to
quickly reduce their demand for grid-supplied electricity. In fact, the emergence of new
concerns about electric system reliability and excessive price volatility – particularly
since the 2000-2001 California energy crisis – has prompted new interest in all forms of
distributed generation. Increased reliance on these resources, it is hoped, can reduce
capacity requirements for central station generators as well as transmission and
distribution infrastructure needs, while improving the competitiveness and stability of
electric markets by allowing a more robust demand response capability to develop on the
customer side. Indeed, the Federal Energy Regulatory Commission (FERC) recently
articulated its interest in promoting demand response capabilities generally in the context
of its proposed Standard Market Design rulemaking:
Allowing demand response infrastructure to satisfy the [resource adequacy]
requirement removes bias toward exclusive reliance on new generation to meet
regional needs. Better demand response to high prices when a shortage condition
approaches will lower demand and reduce the use of high-cost power resources.
Demand response will help ensure reliability, prevent a shortage that could
produce a curtailment, act as a check against market power, and provide a
yardstick for the value that buyers place on supply.3
1
National diesel generator population estimate from Natural Resources Defense Council (NRDC) report
Distributed Resources and Their Emissions: Modeling the Impacts, Greene, Hammerschlag, and Keith,
2001. These figures were provided to NRDC by Power Systems Research (see Footnote 6).
2
National power plant capacity information from EIA’s Annual Energy Review, 2001. Electricity Table
8.7a “Electric Net Summer Capacity: Total 1949-2001.” http://www.eia.doe.gov/emeu/aer/elect.html
3
Though the passage cited refers to “demand response” generally, FERC indicates elsewhere in the
Standard Market Design Notice of Proposed Rulemaking that customer-sited distributed generation can be
viewed as part of demand response, stating for example that: “Distributed generation that is interconnected
with a customer, a load-serving entity, or an energy services company, although it is technically generation
and not demand response, can also be used by a local distributor to reduce the demand that the distribution
1
While increased reliance on distributed generation resources during peak demand periods
can provide obvious benefits in terms of enhancing system reliability and dampening
price volatility and market power opportunities, it raises significant policy concerns from
an environmental standpoint. At present, diesel internal combustion (IC) engines account
for the vast majority of installed distributed generation capacity. When they are operated,
these engines typically emit very high levels of pollutants such as particulate matter (PM)
and nitrogen oxides (NOx), as well as toxic compounds. In fact, per unit of electricity
generated, diesel IC engines emit far more pollution than other distributed generation
technologies (e.g. fuel cells, microturbines, etc.) and far more pollution than most central
station power plants. Figures I-1 and I-2 show emissions rates for a new, uncontrolled
diesel IC engine relative to other generation options. Figure I-1 indicates that NOx
emissions rates per megawatt-hour of electricity generated by a new diesel IC engine are
up to 200 times higher than for a combined cycle natural gas turbine, the technology of
choice for most recent new investments in central-station generator capacity. The fact
that distributed generators are typically located in or nearer to populated areas than large
power plants, and the fact that their emissions are typically vented not through tall
smokestacks, but rather closer to ground level, further amplifies the public health
concerns associated with their use. Finally, such generators are most likely to be operated
during periods of peak electricity demand and high prices, which in many parts of the
country typically occur on summer days when urban air quality tends to be especially
poor.
Figure I-1
NOx Emissions Factors of Diesel Engines Relative to Other Types of Generation4
NOx (lb/MWh)
0.1
Turbine: Large, Combined Cycle, SCR
0.59
Turbine: Large, Simple Cycle
0.4
Turbine: M icrot urbine
1.1
Turbine: Small, Simple
0.6
Turbine: M edium, Simple Cycle
Turbine: ATS, Simple Cycle
0.3
Fuel Cell: Solid Oxide
0.01
0.03
Fuel Cell: Phosphoric Acid
2.2
Engine: Gas fired, Lean Burn
0.5
Engine: Gas f ired, 3-way Catalyst
21.8
Engine: Diesel
4.7
Engine: Diesel, SCR
5.6
U.S. Average Coal Generation
5.1
U.S. Average Fossil Generation
3.4
U.S. Average All Generation
0
5
10
15
20
25
system places on the grid.” The text of the FERC notice is available at:
http://www.ferc.gov/Electric/rto/Mrkt-Strct-comments/discussion_paper.htm.
4
These emissions rates are typical of new units, and do not apply to older existing units. For this reason the
NOx and PM10 emissions factors presented here are smaller than the emissions factors used later in the
report, and smaller than the likely average emissions from the existing population of diesel generators.
Figures I-1 and I-2 are from Joel Bluestein, Emission Rates for New DG Technologies, May 2001 and are
available at: http://www.raponline.org/ProjDocs/DREmsRul/Collfile/DGEmissionsMay2001.pdf.
2
Figure I-2
PM10 Emissions Factors of Diesel Engines Relative to Other Types of Generation4
PM10 (lb/MWh)
0.04
Turbine: Large, Combined Cycle, SCR
0.07
Turbine: Large, Simple Cycle
0.09
Turbine: M icrot urbine
0.08
Turbine: Small, Simple
Turbine: M edium, Simple Cycle
0.07
Turbine: ATS, Simple Cycle
0.07
_
Fuel Cell: Solid Oxide
_
Fuel Cell: Phosphoric Acid
Engine: Gas fired, Lean Burn
0.03
Engine: Gas f ired, 3-way Catalyst
0.03
0.78
Engine: Diesel
0.78
Engine: Diesel, SCR
0.3
U.S. Average Coal Generation
0.27
U.S. Average Fossil Generation
0.19
U.S. Average All Generation
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
State and federal regulators recognize that existing environmental policies will need to be
updated or augmented to ensure that a new generation of cleaner distributed technologies
becomes available in the future and to manage any adverse impacts from the existing
generator population in the transition, especially if market conditions and/or government
policies prompt increased reliance on this population in the near-term. Unfortunately, the
situation is complicated by a shortage of reliable information on the current population of
small, distributed generators. Many existing engines fall below current state permitting
thresholds and/or are permitted for emergency use only. In addition, there are
undoubtedly some units unknown to state authorities that should be permitted but
currently are not. As a result, regulators have had difficulty assessing the potential air
quality impacts of increased use of these resources. Uncertainty about the existing
capacity base of distributed generation also complicates efforts to weigh the costs and
regulatory trade-offs associated with introducing new emissions control or permitting
requirements.
This study aims to begin addressing the current information shortfall, at least in the
Northeast, by developing a more complete inventory of the numbers and types of diesel
IC engines that exist in the eight-state NESCAUM region.5 Two sources of information
were used to develop the inventory: (1) estimates generated by Power Systems Research6
(PSR), a private consultant, using a methodology based on national sales and survey
information combined with census information on numbers and types of business
establishments, and (2) state permitting records. Additional information was gathered
using telephone surveys (also conducted by PSR) for two areas in the Northeast (New
5
Specifically, the states of Connecticut, Maine, Massachusetts, New Hampshire, New Jersey, New York,
Rhode Island and Vermont. All of these states are members of NESCAUM.
6
Power Systems Research (PSR) is a market research company for the engine industry.
3
York City and Fairfield County, Connecticut) where there has been a recent emphasis on
distributed generation because of transmission constraints.
Chapter II provides background and context for these results, including a discussion of
current applications of diesel IC engines in the Northeast and a review of current state
permitting requirements. The results of the inventory analysis and the additional
telephone surveys conducted in New York City and Fairfield County are detailed in
Chapters III and IV of this study. Chapter V presents some estimates of emissions
impacts from the operation of diesel IC engines in the Northeast and summarizes
information available from the New England Independent System Operator on the actual
operation of distributed generators as part of its summer 2001 and 2002 demand response
programs. Chapters VI and VII, prepared by ESI International, Inc., review available
emission control technologies for IC engines and present six case studies involving the
application of these technologies. Finally, Chapter VIII describes current state and federal
activities related to the use and regulation of distributed resources (including stationary
diesel IC engines) and provides some initial policy recommendations.
4
II.
Background and Context
A. Current Role of Electricity-Generating Diesel Engines in the
Northeast
Electricity-generating diesel IC engines can be found at many medium to large
commercial and industrial facilities throughout the Northeast and elsewhere. Often, these
engines are installed to provide emergency power and lighting in outage situations and to
enhance reliability and power quality for sensitive computer and telecommunication
systems. In some cases, on-site generators may be used more routinely to curtail demand
for grid-supplied power during peak electricity use hours (also known as “peakshaving”). Incentives for peak shaving may exist where customers are exposed to realtime electricity prices (which are typically highest during peak demand periods), where
customers have explicitly agreed to curtail their demand for system power when called by
the utility or system operator (usually in exchange for lower rates and/or incentive
payments), or where customers are in a position to avoid or reduce utility demand
charges, which are typically set according to peak usage.
As indicated in the Introduction to this report, emerging concerns about system reliability
and price volatility in deregulated electricity markets have prompted interest in making
greater use of distributed generation capacity. In the long run, this capacity is likely to
include a growing contribution from fuel cells, microturbines, renewable power and other
advanced distributed generation technologies. In the short term, however, diesel IC
engines are likely to remain by far the most ubiquitous distributed generation technology
available. According to the Power Systems Research (PSR) estimates described in later
chapters, the population of diesel IC generators already in place in the Northeast states
may number well over 30,000 units with a combined capacity in excess of 10 GW.7 Of
these engines, the great majority (80%) is estimated to be installed for emergency use; the
remaining 19% is operated for peak-shaving purposes, and the last 1% is intended for
baseload operation. Engines designated for emergency use only are generally limited by
state regulations to operate only during bona-fide “emergency” (i.e. outage or imminent
outage) situations and for a specified maximum number of hours of operation per year (a
typical figure is 500 hours). Given the infrequency of actual outage events in recent
years, most of these engines should have operated well below state-imposed limits – in
many cases less than the 50 hours or so of annual operation required for maintenance and
testing. Depending on permit limits and market conditions, peak shaving engines – by
contrast – may operate considerably more hours per year (typically anywhere from 200 to
700 hours, annually). Because electricity demand peaks in the Northeast on hot summer
days, this is the time when all types of small engines are most likely to operate, whether
for emergency “back-up” purposes or in response to high grid prices and/or supply
shortfalls.
7
By comparison, total grid-connected generating capacity totaled some 86 GW in 2000 for the NESCAUM
region (capacity data from EIA: http://www.eia.doe.gov/cneaf/electricity/ipp/ipp_sum.html and
http://www.eia.doe.gov/cneaf/electricity/ipp/ipp_sum2.html).
5
At present, both the New York and New England Independent System Operators (ISOs)
have introduced formal programs aimed at improving system reliability.8 These programs
enlist customers to lower their demand for grid-supplied electricity at times of high prices
and/or when demand threatens to overwhelm system capacity. Customers can reduce
demand by curtailing load (e.g. shutting down equipment), increasing on-site generation
or some combination of both. Under the ISO programs introduced to date, customers can
receive incentives, including capacity payments, for being available to reduce load on
short notice. Subject to certain conditions, some customers may also be eligible to bid
their demand resources into day-ahead wholesale electricity markets in the same way that
generation owners bid in supply resources.9 Because participants in these programs must
be registered with the ISO, it is generally possible to collect some information on
individual on-site generators operated as part of the customer’s demand response
capability.10 In the future, however, other types of customer-initiated demand response
may be more difficult to monitor. For example, customers (particularly larger commercial
and industrial customers) that opt to purchase electricity based on real-time spot market
prices might face significant incentives to reduce their consumption of grid-supplied
electricity during high price periods, even without participating in an ISO or utilitysponsored program. If reducing demand in these instances involved the utilization of onsite generators, neither the ISO nor state regulators would necessarily be aware of it.
The policy issues associated with a potentially significant increase in the use of
distributed generators are explored in more detail in Chapter VIII of this report, following
a discussion of population estimates, survey results, potential emissions impacts, control
technologies and case studies in Chapters III, IV, V, VI and VII, respectively. To provide
context for the information presented in these chapters, it is useful to begin by reviewing
current state permitting requirements and regulations as they pertain to the use of powergenerating diesel IC engines in the Northeast states.
8
Note that New Jersey, the southern-most NESCAUM state, is served by PJM which is the regional
transmission organization that coordinates the wholesale electricity market for all or parts of Delaware,
Maryland, New Jersey, Pennsylvania, Virginia, West Virginia and the District of Columbia. The acronym
‘PJM’ comes from the original Pennsylvania-New Jersey-Maryland power pool. PJM is the largest of the
three ISOs serving northeastern states and is scheduled to become even larger in the future as additional
areas to the south and west become members.
9
Generally, Northeast states have prohibited engines permitted for emergency use from participating in
ISO-sponsored demand response programs. In particularly transmission-constrained areas such as
southwest Connecticut and New York City, emergency generators have been allowed to operate – subject
to certain limitations – in the ISO’s emergency response programs, which are designed to avert imminent
shortfalls and are called only when the system is strained to the point of necessitating voltage reductions.
So far, emergency generators have generally not been eligible to participate in purely economic or so-called
“price-response” programs which are triggered by high prices, rather than capacity shortfalls. See further
discussion in Chapters V and VIII.
10
In fact, a recent recommendation of the New England Demand Response Initiative described in Chapter
VIII involves information collection requirements, both for participants in ISO-sponsored demand response
programs and on the part of the ISO managing these programs.
6
B. Summary of Current State Permitting Requirements for Diesel IC
Engines
Distributed generators, and stationary internal combustion engines in general, are for the
most part regulated and permitted at the state and local level. Some large engines fall
under federal regulations, which include the New Source Review (NSR) program for
permitting sources before construction, and the Title V program for permitting sources in
operation. These engines are generally far larger than any considered in this report. The
NSR program includes a major and minor source component. The minor source
component is applied in some states to non-emergency engines that have the potential to
exceed certain pollution thresholds (applicable thresholds are lower for ozone nonattainment areas). Because there are no federal requirements for many smaller engines,
states over the years have generally developed their own permitting requirements for this
class of emissions sources.
Table II-1 summarizes the different permitting requirements applicable to electricity
generating engines in the eight NESCAUM states. As indicated by Table II-1, most states
make a distinction between emergency and non-emergency engines. Emergency engines
are often exempt from emissions limits or control technology standards, however their
operation is strictly limited to certain situations and maximum numbers of hours. In many
states, owners of emergency engines need not obtain individual permits but can avail
themselves of a permit-by-rule option which authorizes operation provided the owner
complies with a prescribed set of restrictions and requirements – typically concerning
maximum rated heat input, fuel type, fuel consumption and record keeping. By contrast,
non-emergency engines are generally regulated down to smaller sizes and under more
stringent emissions control requirements.
Though most state permitting programs share several common features, the specific
requirements and permitting thresholds applicable to distributed generators vary widely
from state to state. Many states regulate engines on the basis of fuel input or heat rate
(e.g. million British thermal units per hour: MMBtu/hr); others use engine power output
(e.g. horsepower, brake horsepower or kilowatt) to determine the applicability of
different regulatory requirements. Converting heat input to power output is
straightforward, provided the efficiency of the engine is known. Throughout this report,
the generic efficiency assumed for stationary generator engines is 33%.11 This
assumption yields the following conversion factors:
1 MMBtu/hr = 100 kW = 0.1 MW = 134 hp
11
Engine efficiency is measured as the ratio of useful energy output to fuel energy input. In the case of an
electric generator, an efficiency of 33% indicates that one-third of the energy content of the input fuel is
converted to electrical power, while the remaining two-thirds is lost (usually in the form of waste heat).
Combined heat and power systems can achieve much higher efficiencies because they capture this
otherwise wasted heat to supply useful thermal energy (typically in the form of steam) as well as electricity.
7
Table II-1
Summary of State Permitting Requirements for Distributed Generators
State
Non-Emergency Enginesa
Threshold
Emergency Engines
Requirementsb
Threshold
Restrictions
Demand Responsed
500 hrs/yr and
no PRP; EDRP in SW
CT: permit-by-rule
maximum of 5 TPY
CT for add'l 300 hrs/yr,
SW CT: 50 hp (37
NOx, 5 TPY CO, 3 TPY
kW)
only nat. gas or ULSDe
PM, and 3 TPY SO2
CT
PTE 15 TPYc of any
criteria pollutant
ME
5 MMBtu/hr (approx.
SCR over 20 TPY NOx, 0.5 MMBtu/hr
500 kW), 0.5 MMBtu/hr BACT case-by-case, on- (approximately 50
if at major source
road diesel
kW)
500 hrs/yr
no additional
restrictions
MA
3 MMBtu/hr (approx.
300 kW), smaller if at
facility with other
permitted engines
3 MMBtu/hr permitby-rule, over 10
MMBtu/hr case-bycase BACT
300 hrs/yr, cannot
create a "condition of air
pollution," must have a
noise muffler
no PRP; may run once
ISO has called for
voltage reductions (OP4 step 12 or 14)
no threshold
500 hrs/yr, limit sulfur
content of diesel, and
limit emissions
neither type of DR for
emergency engines
NH
NJ
BACT, LAER based on
emissions
case-by-case BACT
1.5 MMBtu/hr (150 kW)
diesel, 10 MMBtu/hr (1 over 400 kW may
MW) natural gas, PTE require RACT
25 TPYc NOx
BACT for new/modified;
existing diesel engines
1 MMBtu/hr
require 8g/bhp-hr NOx
(approximately 100 kW) (being revised to 2.3
g/bhp-hr)
1 MMBtu/hr
no control if PTE NOx is neither type of DR for
(approximately 100
less than 25 TPY
emergency engines
kW)
NY
NY: 300 kW, 160 kW if diesel engines are not
NY: no threshold, NY: 500 hrs/yr, no
non-att., NYC: 280 kW, allowed to participate in NYC: over 280 kW permits, NYC: register
33 kW if diesel
must register
but no restrictions
PRPd
RI
500 kW diesel,
1 MW natural gas
VT
450 hp (337 kW)
BACT based on
emissions
must meet EPA's nonroad standards
no threshold
500 hrs/yr, 0.3% sulfur
diesel fuel
no threshold
200 hrs/yr
no PRP; EDRP less
than 200 hrs/yr, 30 ppm
sulfur diesel fuel
required
neither type of DR for
emergency engines
neither type of DR for
emergency engines
a
non-emergency engines are not restricted from participating in demand response programs, except as noted in NY
abbreviations: BACT=Best Available Control Technology; MACT=Maximum Achievable Control Technology; RACT=
Reasonably Available Control Technology; LAER=Lowest Achievable Emission Rate; SCR=Selective Catalytic Reduction;
SOTA=State of the Art
b
c
PTE=potential to emit, and TPY=tons per year
demand response (DR) programs include the emergency demand response program (EDRP) which is called by the ISO in the
event of an imminent capacity shortfall, and the price response program (PRP) in which customers respond to high prices
d
e
ultra-low sulfur diesel fuel
1. Connecticut
Connecticut’s permitting program currently provides four tracks for regulating IC
engines. The first track involves the application of NSR minor source requirements to
individual units. Prior to December 1989, NSR applied to any combustion source with
fuel input greater than or equal to 5 MMBtu/hr heat input, approximately 500 kW output,
or a potential to emit in excess of 8 tons per year of any criteria air pollutant. After 1989,
NSR applicability thresholds were lowered to include any source with the potential to
emit 5 tons per year or more of any criteria pollutant. Due to the large number of small
engines being reviewed, changes were made to the NSR requirements in 2002. Currently,
a new or modified engine with the potential to emit 15 tons per year or more of any air
pollutant requires an individual permit. Engines permitted under NSR are subject to Best
8
Available Control Technology (BACT) and Lowest Achievable Emission Rate (LAER)
requirements as necessary, based on emissions. There may also be “grandfathered”
sources that are registered but unrestricted.
A second compliance option, introduced in 1996, allows emergency engines to register
with the state under a General Permit for an Emergency Engine (GPEE). This general
permit limits annual operation of emergency units to a maximum of 500 hours each year
and limits combined emissions of nitrogen oxides (NOx), particulate matter (PM) and
carbon monoxide (CO) to a maximum of 5 tons per year, among other restrictions. In
2002, Connecticut introduced a third compliance option known as the General Permit for
Distributed Generation (GPDG). This option is geographically restricted to engines over
50 hp (37 kW) located in the 51 towns that comprise the southwest Connecticut “load
pocket” and it expires at the end of 2003. The GPDG allows engines to operate when
called upon by the ISO under the emergency demand response program for a maximum
of 300 hours per year, provided they run on natural gas or ultra-low sulfur diesel fuel.12 A
final compliance option, also introduced in March 2002, is Connecticut’s permit-by-rule
exemption, which allows emergency engines to operate subject to certain restrictions on
maximum rated heat input, fuel type and fuel consumption. In addition, the exemption
requires the owner of the engine to be responsible for record keeping. Connecticut’s
permit-by-rule exemption does not allow for the operation of emergency engines in any
type of demand response program. In order for an emergency engine to participate in ISO
New England’s emergency demand response program, the engine must be permitted
under the GPDG (and thus be located in southwest CT). Engines permitted for nonemergency use in Connecticut are allowed to participate in both price response and
emergency demand response programs, regardless of location.
2. Maine
In Maine, owners of engines with heat input greater than 5 MMBtu/hr (approximately
500 kW output) must obtain a permit. Permits are also required for smaller engines
(down to a heat rate input of 0.5 MMBtu/hr, or approximately 50 kW) if they are located
at a facility with a combined heat input of 5 MMBtu/hr or more. Finally, facilities with
operation-specific air permits must obtain permits for any on-site engines larger than 0.5
MMBtu/hr. Non-emergency engines are required to use on-road diesel fuel and are
required to install selective catalytic reduction (SCR) technology for NOx control if their
potential annual NOx emissions exceed 20 tons. Emergency engines larger than 0.5
MMBtu/hr require a permit, and are restricted to no more than 500 hours of operation
each year. There are no additional restrictions preventing engines from participating in
demand response programs.
12
In the emergency demand response program, the NE-ISO can call on participating distributed generators
to help avert an imminent power shortage. Generally, the ISO must have called for an actual system voltage
reduction before emergency units are eligible to respond under this type of program. By contrast, under a
price response program, distributed generators can operate whenever prices are high, even if a power
shortage is not necessarily imminent. Since high price periods occur more frequently than voltage
reduction/imminent shortfall situations, which in turn occur more frequently than actual emergency (i.e.
black-out) situations, generators can be expected to operate with varying degrees of frequency depending
on their eligibility to participate in these different types of programs.
9
3. Massachusetts
In Massachusetts, permits or “plan approval” must be obtained for all engines (including
emergency engines) with heat input greater than 10 MMBtu/hr (approximately 1 MW
output). A lower permitting threshold of 3 MMBtu/hr (approximately 300 kW) applies to
any engines intended for use in non-emergency situations. Facilities with a combined
heat input greater than 10 MMBtu/hr must file a statement of emissions at least every
three years. Permits for non-emergency engines are reviewed on a case-by-case basis and
generally require BACT-level emissions controls. Emergency engines between 3 and 10
MMBtu/hr heat input may operate under permit-by-rule, but are restricted for use only
during emergencies and for a total of no more than 300 hours per year. The
Massachusetts Department of Environmental Protection (MA DEP) has recently
indicated that emergency generators may be allowed to operate to avert imminent outage,
but only after the ISO has called for voltage reductions (i.e. corresponding to Step 12 or
14 of ISO-NE’s Operating Procedure 4).13 Emergency engines in Massachusetts are
prohibited from operating in price response programs.
4. New Hampshire
Engines powered by liquid fuel (i.e. diesel) require a permit in New Hampshire if they
have a heat rate input of 1.5 MMBtu/hr (approximately 150 kW output) or greater. A
higher size threshold of 10 MMBtu/hr (1 MW output) applies to engines that operate on
gaseous fuel. Additionally, any type of engine requires a permit if it emits more than 25
tons per year of NOx. Non-emergency engines larger than 550 hp (approximately 400
kW) may be required to implement Reasonably Available Control Technology (RACT).
Emergency engines can obtain a general permit for emergency exemption, which limits
operation to a maximum of 500 hours per year, requires use of low-sulfur fuel and
restricts total emissions. Engines with an emergency exemption may operate only in bona
fide emergency situations (i.e. they cannot participate in emergency demand response or
price response programs).
5. New Jersey
In New Jersey, all new or modified engines with a heat rate input greater than 1
MMBtu/hr (equivalent to about 100 kW output) require a permit. In addition, any new or
modified engine with the potential to emit more than 5 tons per year of any criteria
pollutants must meet “state of the art” (SOTA) control technology requirements. The
recently published (May 2003) applicable SOTA performance standards for new or
modified engines are 0.15 g/bhp-hr for NOx, 0.5 g/bhp-hr for CO and 0.15 g/bhp-hr for
volatile organic compounds (VOC). In addition, ammonia slip is limited to 10 ppmvd @
15% O2. For liquid fuel firing, the particulate limit is 0.02 g/bhp-hr and the sulfur limit is
30 ppm. Meanwhile, existing engines larger than 500 bhp (approximately 375 kW) must
also comply with minimum emissions performance requirements, specifically: (1) a rich
13
Letter of James C. Colman, Assistant Commissioner, Bureau of Waste Prevention, MA DEP to Steve
Whitley, Chief Operating Officer, ISO-NE ( May 28, 2002).
10
burn NOx emissions limit of 1.5 g/bhp-hr and a lean burn NOx emissions limit of 2.5
g/bhp-hr for gaseous fuels; (2) a NOx emissions limit of 8.0 g/bhp-hr for liquid fuels; and
(3) a CO emissions limit (on all engines) of 500 ppmvd at 15% O2. Emergency engines
are exempt from NOx control requirements provided they operate no more than 500
hours per year and only during actual emergencies or for maintenance purposes; and
provided their potential to emit NOx is below 25 tons per year.
6. New York
The New York State Department of Environmental Conservation (NY DEC) has
established a permitting threshold for IC engines in ozone attainment areas of 400 bhp
(approximately 300 kW). In ozone non-attainment areas (New York City, Long Island,
and the lower Hudson Valley) a lower permitting threshold of 225 bhp (160 kW) applies.
Emergency generators are exempt from permit requirements but are limited to a
maximum of 500 hours per year of operation and may operate only when usual sources of
heat and power are not available and during fire emergencies. To verify compliance,
emergency generators are required to maintain records of operation on-site for five years.
Emergency diesel generators may participate in the emergency demand response program
called by the ISO to avert outage situations, but must use 30 ppm ultra-low sulfur fuel
and are restricted to operating a maximum of 200 hours per year in the program (as part
of their overall operational limit of 500 hours per year). Finally, the state of New York
has banned all emergency generators and all diesel generators from participating in the
price response program.
NY DEC issues three types of permits: (1) “Registration certificates” with a “cap-byrule” which restricts actual NOx emissions in ozone non-attainment areas to no more than
12.5 tons per year and NOx emissions in other areas to no more than 50 tons per year; (2)
state facility permits for facilities that do not qualify for a registration certificate, but
whose potential to emit is lower than the threshold for Title V permits; and (3) Title V
permits, if the potential to emit is higher than Title V thresholds.14 Additional permitting
requirements are enforced by the New York City Department of Environmental
Protection (as distinct from the NY State DEC) for units located in New York City. In
New York City, all diesel engines over 33 kW must – at a minimum – be registered.
Engines over 2.8 MMBtu/hr (280 kW) must obtain a work permit unless they are
emergency engines, in which case they must simply register. Emergency engines are not
restricted in terms of the number of hours they may operate in emergency situations. In
addition, they may participate in emergency demand response programs, but may not be
used for peak or baseload generation and must comply with state regulations.
14
Title V permits are issued under the 1990 Clean Air Act Amendments to all existing major sources (as
distinct from new or modified sources, which are regulated under NSR). The emissions thresholds required
to classify a source as “major” depend on the attainment status of the area for different pollutants. For areas
in attainment, the major source threshold is usually a PTE of 100 TPY; in moderate non-attainment areas,
the threshold is lower at a PTE of 50 TPY, and in severe non-attainment areas the threshold is a PTE of 25
TPY. Note that the lower thresholds for non-attainment areas apply only to those pollutants (or their
precursors) for which the area is in non-attainment.
11
7. Rhode Island
Diesel engines in Rhode Island must obtain a pre-construction permit if their heat input is
greater than 5 MMBtu/hr (approximately 500 kW). If operating on natural gas, an engine
over 10 MMBtu (approximately 1 MW) requires a permit. Additionally, smaller engines
(down to a size threshold of 1 MMBtu/hr or 100 kW) must be included in facility
operating permits if located at a major source facility. Engines for emergency use are
allowed to operate for no more than 500 hours per year and only during power outages.
In addition, diesel fuel used to operate an emergency engine in Rhode Island must have
sulfur content no higher than 0.3%.
8. Vermont
The permitting threshold for generator engines in Vermont is 450 bhp (337 kW). Engines
above this size threshold that were installed after June 1999 must meet emissions
standards comparable to federal requirements for non-road sources. Engines larger than
200 bhp (150 kW) that are located at an otherwise permitted facility must be covered by
amendments to the permit. Emergency engines can operate a maximum of 200 hours per
year, during emergency situations only, and are not allowed to participate in emergency
demand response or price response programs.
12
III. Northeast States Inventory
A. Introduction
This chapter describes the results of an effort to inventory the existing population of
distributed generator engines in the eight NESCAUM states. It combines information
gathered from state environmental agencies with population estimates generated by
Power Systems Research (PSR) using a methodology developed from engine sales data
and field surveys. Together, these two sources of information provide estimates of both
the permitted and the total engine population in the NESCAUM region, including the
number of engines in each state, their capacity and whether they are covered by state
permitting requirements. The results provide some indication of the extent to which
current state permitting programs capture the population of existing generators.
This chapter begins with a description of the methodology and data sources used to
compile the Northeast distributed generation inventory. Later sections summarize the
results for the region as a whole and then for individual states, each of which has
different permitting requirements and record-keeping practices.
B. Methodology
1.
PSR Population Estimates
PSR developed estimates of the population of distributed generation engines in the
NESCAUM region using their Partslink database.15 The database uses a mathematical
model, developed by PSR over the last 23 years, to estimate the number of “engine
powered products” in service in the United States based on actual sales data and survey
results. Key inputs to the model include: (1) a continuous record of shipments from U.S.
factories that manufacture engines, as well as records of imports from foreign suppliers;
(2) a record of exports from the U.S. of this type of equipment; and (3) an attrition curve
for estimating the retirement of “engine powered products.”
To translate data on the number of engines sold into an estimate of engines currently in
use, a number of assumptions must be made about equipment turnover or attrition, taking
into account factors such as the average life of an engine, hours of use each year and
typical load factors for different engine applications. PSR continually checks and updates
benchmarks for these characteristics by conducting surveys of equipment owners. PSR
surveys 100 owners for each engine model and application each year to obtain
information on the age and application of the equipment, as well as cumulative lifetime
hours of operation, annual hours of operation and fuel consumption. Information on
average total hours of operation for different models (and at different load factors) is
typically available from engine manufacturers and other sources, but fuel use – which is
15
A more detailed description of the methodology, provided by PSR, is included as Appendix A.
13
directly related to load factor – can be used as an alternate indicator to verify these
assumptions. Data on engine characteristics, such as horsepower, speed, cylinders, weight
and displacement, are then combined with the survey results to calculate a projected life
span for every engine in the database. This estimated life span, expressed in horsepowerhours, is assumed to represent the statistical mean for each engine type and application
and a normal distribution is used to describe all retirements.
The Partslink model includes detailed information on engines for many different uses.
For the NESCAUM inventory, PSR provided information on stationary and portable
electricity-producing generators powered by diesel and natural gas in the eight Northeast
states. As noted previously, the inventory subsequently compiled by NESCAUM using
PSR’s estimates focuses only on stationary diesel generators because this is the
population most likely to be captured in state permit records.16 In addition, PSR provided
estimates of the geographic distribution of different engine types based on profile norms
that describe the probability that a certain type of business or consumer will own a
specific type of equipment. The profile norms were combined with the Census Bureau’s
County Business Patterns (year 2000) statistics to allocate types and numbers of engines
based on the distribution of businesses in each state. A spreadsheet for estimating the
number of engines in any state or other geographic area by industry is provided in
Appendix C of this report.
In addition to estimating numbers of engines, PSR developed estimates of the distribution
of engine capacities and applications in each state. Engine applications, described in more
detail in Chapter II, are divided into three categories – emergency, baseload and peak
shaving – depending on typical hours of use and purpose. Emergency engines are used to
power a facility or building in the event of a power outage. Absent such events, they may
operate as few as 50 hours per year (for testing and maintenance). Even when
emergencies occur, their cumulative operating hours are unlikely to exceed 100-150
hours per year. Baseload engines are used as a primary power source, and can run as
infrequently as 700 hrs/yr to nearly full-time (up to 8760 hrs/yr). Engines used for peak
shaving are defined more by their purpose than hours of use, but they generally run fewer
than 700 hrs/yr. Peak shavers are usually operated at times of peak demand for gridsupplied electricity to avoid high electricity prices and/or demand charges, which are
typically based on peak usage and can be quite expensive.
While PSR’s estimation methodology represents one of the few sources of information
available on the total population of distributed generators, it is inherently inaccurate.
Because multiple assumptions are involved in generating the estimates, it is difficult to
know how much confidence can be placed in the results. Comparison with state permit
records (as described in the remainder of this chapter) reveals both that the discrepancies
are significant and that there is no particular pattern to the variance between states’
permit data and PSR’s population estimates for different engine size categories. This
suggests either that the PSR estimates are highly inexact, or that available state permit
records are incomplete, or some combination of both.
16
A chart showing the estimated population of natural gas generator engines in the eight NESCAUM states
is included as Appendix B of this report.
14
2. State Permit Data
NESCAUM requested a detailed list of permitted engines from the air quality bureau of
each state’s environmental agency. Typically, the state lists included details such as the
size of the engine (in kW capacity, heat input or brake horsepower), application, fuel,
emissions and control technology. Some states were not able to provide details for each
listed engine and it was sometimes necessary to extrapolate the size distribution of
engines when size information was not available. The extrapolations used were different
in each case, and are described in detail in the state summaries below.
For the most part, state permit data were provided electronically. For some states
(primarily MA and NJ) this meant the information did not include engines permitted prior
to the mid-1990s, when many states switched over to electronic record keeping. Other
states have transferred all of their permits to electronic files and were able to provide
information on older permits as well. For Rhode Island and Vermont, NESCAUM
actually reviewed paper permits.17 This was feasible because of the relatively smaller
number of permitted engines in these two states.
For purposes of this report, NESCAUM’s primary interest was diesel-powered internal
combustion engines used to generate electricity. State permit records included a number
of types of equipment that do not fall in this category, such as diesel engines used for
primarily mechanical purposes (e.g. stone cutters, wood chippers, and snow makers), as
well as small engines used as fire pumps, air compressors, and water pumps. The latter
types of equipment are typically under 250 kW and are usually permitted as emergency
engines subject to annual caps on hours of operation. State permit data on mechanical IC
engines are summarized separately in the tables that follow, because PSR’s estimates
were designed to cover only electricity-generating engines. Nevertheless, both types of
engines have similar level of emissions. In addition to mechanical engines, state permit
records included some non-internal combustion equipment, such as gas turbines and
boilers, which were also removed for purposes of comparison with PSR's population
estimates.
C. Northeast Distributed Generation Inventory Results
1. NESCAUM Region Totals
Table III-1 shows the results of PSR’s estimation methodology for the NESCAUM states
as a region. Engine totals are sorted by size, capacity and type of application.
17
In the case of Rhode Island, it was necessary to review the state’s paper permits to get information on
engine size. In the case of Vermont, only paper permits were available.
15
Table III-1
PSR Estimates of Diesel Engines in the NESCAUM Region by Number and Capacity
Number Totals Emergency Peak Baseload Total
25-50 kW
1,768
0
0
1,768
50-100 kW
5,798
1,375
107
7,280
100-250 kW
9,226
2,236
95
11,557
5,918
1,231
7
250-500 kW
7,156
500-750 kW
1,296
316
47
1,659
750-1000 kW
1,164
292
51
1,507
1000-1500 kW
641
677
39
1,357
1,073
284
37
1500+ kW
1,394
Total
26,884
6,411
383
33,678
Capacity Totals (MW) Emergency Peak Baseload Total
25-50 kW
59
0
0
59
50-100 kW
462
114
9
584
100-250 kW
1,564
371
14
1,949
2,126
443
3
250-500 kW
2,572
500-750 kW
801
196
29
1,026
750-1000 kW
921
230
40
1,191
1000-1500 kW
769
837
48
1,654
2,053
615
68
1500+ kW
2,736
Total
8,756
2,805
211
11,772
Overall, PSR estimates that a total of 33,678 diesel IC engines, with a total of capacity of
11,772 MW, are currently installed in the eight NESCAUM states. Of this population, an
estimated 80% of the engines (accounting for 74% of total capacity) are designated for
emergency use.
Information on the electricity-generating IC engine population (distinct from the
mechanical IC engine population) from the permit records of all eight NESCAUM states
is summarized in Table III-2, below. As noted previously, state permit records differed in
the amount of information and detail available. In Connecticut, Massachusetts and New
Jersey, size or capacity information was available for only some of the permitted engines.
In these cases, the size distribution for the subset of known engines was assumed to be
representative of the total population of permitted engines. Further detail on the
information available from different states is provided in the state summaries that follow.
Table III-2
Number of Electricity-Generating Engines in NESCAUM State Permit Records
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
CT
112
208
411
321
273
144
153
99
1,721
ME
2
78
184
158
64
28
36
28
578
MA
11
13
278
156
138
73
160
275
1,104
NH
1
2
65
126
71
39
47
9
360
NJ
4
120
1,432
1,247
927
837
698
558
5,823
NY
26
93
337
410
272
201
175
148
1,662
RI
0
0
4
1
20
11
11
25
72
VT
0
9
18
17
7
2
10
3
66
TOTAL
156
523
2,729
2,436
1,772
1,335
1,290
1,145
11,386
Table III-3 compares PSR’s total estimated population figures with the permit records
available from each NESCAUM state for only the electricity-generating IC engines. With
the exception of Maine, PSR estimates indicate a significantly larger engine population in
place than state permit records alone would indicate. For the other states, permits are on
file for anywhere from 11% to 69% of the total engine population estimated by PSR and
in four of the states (MA, NY, RI, VT) the figure is below 25%. This is not surprising for
several reasons. First and most obvious, is the fact that many engines in each state fall
16
below the size thresholds of current permitting requirements, which typically range from
75 kW to 500 kW, depending on the state. While some small engines fall within “facility
clauses” which require smaller engines to register if they are located at a larger facility,
many others simply do not require a permit. In several states a large number of engines
may also operate under emergency exemptions or permit-by-rule provisions, in which
case the state would not have records for individual units. Another explanation is that
some older permits still filed on paper may not have been included in the electronic
records most states provided to NESCAUM. In addition, there are undoubtedly some
engines that should have permits, but whose owners do not obtain them, either because
they are unaware of permitting requirements or because they wish to avoid the fees,
hassle or restrictions associated with obtaining a permit. Finally, the PSR estimates
themselves – as indicated previously – are subject to significant uncertainties and errors
in approximation.18
Table III-3
Comparison of PSR Estimated Generators and State Permit Records
DATA SOURCE
PSR Estimated
Generators
Generators in State
Permit Records
% of PSR Estimates in
Permitting Records
CT
ME
MA
NH
3,223
560
5,027
743
1,721
578
1,104
53%
103%
22%
NJ
NY
RI
VT
Total
8,415 15,037
363
310
33,678
360
5,823
1,662
72
66
11,386
48%
69%
11%
20%
21%
34%
As indicated previously, engines used for mechanical power – wherever these were
distinguishable from the general engine population – were treated distinctly in the
development of this preliminary regional inventory. In particular, mechanical engines
were not compared to the population estimates generated by PSR (as these estimates were
designed to include only IC engines used to generate electricity) and are not included in
Table III-3 above. Instead, Table III-4 below summarizes information on mechanical IC
engines included in the state permit data obtained by NESCAUM for the region as a
whole. These mechanical engines represent 15% of the total population of permitted
engines in the region, and 12% of the total engine capacity estimated for the NESCAUM
region.
18
In fact, the discrepancy between the PSR figures and state permit records may in fact be greater than
indicated by these tables, because the state totals in Table III-2 may include some non-diesel engines. In
cases where a state permit did not identify fuel source, the engine was assumed to operate on diesel. This
likely led to some overstatement of the numbers of diesel engines in state permit records.
17
Table III-4
Mechanical Engines in NESCAUM State Permit Records
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
CT
5
31
91
40
16
7
2
13
205
ME
0
11
51
16
5
2
0
0
85
MA
0
2
33
19
12
2
5
12
85
NH
1
0
12
5
4
0
0
0
22
NJ
0
32
314
374
97
0
20
34
871
NY
44
37
43
21
4
2
1
7
159
RI
0
0
0
0
11
0
0
3
14
VT
1
14
31
73
9
14
8
0
150
Total
51
127
575
548
158
27
36
69
1,591
2. Connecticut Inventory
Table III-5 summarizes the engine population and size distribution estimated by PSR for
the state of Connecticut. According to these estimates, over 81% of diesel engines and
76% of total engine capacity in Connecticut are designated for emergency use.
Table III-5
PSR Estimates of Diesel Engines in Connecticut by Number and Capacity
Number Totals Emergency Peak Baseload Total
168
0
0
25-50 kW
168
50-100 kW
555
123
2
680
100-250 kW
883
202
3
1,088
250-500 kW
563
118
0
681
500-750 kW
127
28
1
156
750-1000 kW
112
27
3
142
63
64
2
1000-1500 kW
129
1500+ kW
143
33
3
179
Total
2,614
595
14
3,223
Capacity Totals (MW) Emergency Peak Baseload
6
0
0
25-50 kW
50-100 kW
45
10
0
100-250 kW
149
34
0
250-500 kW
202
43
0
500-750 kW
78
17
1
750-1000 kW
89
21
2
76
79
3
1000-1500 kW
1500+ kW
275
70
5
Total
919
275
11
Total
6
55
183
245
96
112
158
350
1,205
Connecticut provided NESCAUM with two databases of information on specific IC
engines. The first is Connecticut’s emergency engine database, which includes 1,085
electricity-generating engines whose owners applied for the state’s General Permit for
Emergency Engines from 1996 through early 2002. The emergency engine database
provided specific size or capacity information on 845 of the engines listed; the size
distribution of the remaining 240 engines for which this information was not available
was assumed to be the same for purposes of this inventory. A second state database
provides information on 636 non-emergency engines used to produce electricity,
including heat input and fuel type. Some of the engines in this database are not permitted,
but their size and location are noted because they are installed at a major source facility.
The size distribution for all 1,721 generators listed in the two Connecticut databases is
shown in the left-hand side of Table III-6. The right-hand side of Table III-6 summarizes
information on an additional 205 engines in the Connecticut database that are used in
mechanical applications (as opposed to generating electricity).
18
Table III-6
Connecticut Permit Records for Electricity Generators and Mechanical Engines
Electricity
Generators
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Number of
Engines
112
208
411
321
273
144
153
99
1,721
Capacity
Totals (MW)
4
13
60
111
158
118
174
219
857
Mechanical
Engines
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Number of
Engines
5
31
91
40
16
7
2
13
205
Capacity
Totals (MW)
<1
2
12
12
9
6
2
29
74
Table III-7 summarizes the agreement between PSR’s estimates and state permitting
records for different size categories of engines. The red line in Table III-7 and in
subsequent tables comparing PSR estimates to state records shows the operative
permitting threshold for most engines in each state. In theory, there should be much
greater agreement between PSR’s estimates and state records in the size categories over
applicable permitting thresholds. Therefore, the last two rows in Table III-7 show overall
agreement for all size categories, as well agreement for just those categories over the
applicable permitting threshold. The fact that substantial discrepancies remain in the
larger size categories may nevertheless be explained by a number of factors, including the
existence of permit-by-rule provisions, different permitting thresholds for emergency and
non-emergency engines, incomplete state information and uncertainties in PSR’s
estimation methodology.
Table III-7
Connecticut Engine Data Comparison by Number and Capacity
PSR
25-50 kW
50-75 kW
75-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Total above permit size
160
227
461
1,088
681
156
142
129
179
3,223
2,836
TOTAL ENGINES
% State/PSR
State
Agreement
112
70%
153
67%
55
12%
411
38%
321
47%
273
175%
144
101%
153
119%
99
55%
1,721
53%
1,456
51%
19
TOTAL CAPACITY (MW)
% State/PSR
PSR
State
Agreement
5
4
79%
15
9
58%
41
4
11%
183
60
33%
245
111
45%
96
158
163%
112
118
106%
158
174
110%
350
219
63%
1,205
857
71%
1,185
844
71%
3. Maine Inventory
Table III-8 summarizes PSR’s engine population and size distribution estimates for the
state of Maine. According to these estimates, emergency engines account for over 86% of
all diesel engines and 69% of total engine capacity in Maine.
Table III-8
PSR Estimates of Diesel Engines in Maine by Number and Capacity
Number Totals Emergency Peak Baseload Total
25-50 kW
29
0
0
29
50-100 kW
120
107
13
0
100-250 kW
175
159
16
0
250-500 kW
117
111
6
0
500-750 kW
20
20
0
0
750-1000 kW
22
3
0
25
1000-1500 kW
9
4
5
0
1500+ kW
65
34
27
4
Total
486
70
4
560
Capacity Totals (MW) Emergency Peak Baseload
25-50 kW
0
0
0
50-100 kW
9
1
0
100-250 kW
27
3
0
250-500 kW
41
2
0
500-750 kW
12
0
0
750-1000 kW
17
2
0
1000-1500 kW
5
6
0
1500+ kW
64
57
8
Total
175
72
8
Total
0
10
30
43
12
20
11
129
255
The Maine Department of Environmental Protection (ME DEP) provided NESCAUM
with a list of all IC engines currently permitted in the state, including emergency engines.
The state's permitting guidelines indicate that Maine’s permits should capture all engines
larger than 500 kW and many smaller ones as well. Maine currently allows engines
smaller than 50 kW to operate without a permit or any other restriction on hours of
operation or emissions. Non-emergency engines between 50 kW and 500 kW may also
fall outside permitting requirements, unless they are located at a facility with multiple
emission sources. In addition to the engines shown on the left-hand side of Table III-9,
Maine’s database included 85 engines that were used in mechanical applications. These
engines, which are summarized on the right-hand side Table III-9, were mostly
emergency fire pumps, drives and wood chippers.
Table III-9
Maine Permit Records for Electricity Generators and Mechanical Engines
Electricity
Generators
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Number of
Engines
2
78
184
158
64
28
36
28
578
Capacity
Totals (MW)
<1
6
32
56
40
23
45
60
262
Mechanical
Engines
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Number of
Engines
0
11
51
16
5
2
0
0
85
Capacity
Totals (MW)
0
1
8
5
2
2
0
0
19
Table III-10 compares PSR’s engine population estimates for Maine with the information
contained in state permit records. As noted earlier in this Chapter, Maine is the only
NESCAUM state for which PSR’s estimates under predict the total number of engines
20
identified in state permit records. This is not the case in all size categories, however, as
indicated in Table III-9. State permit records indicate a greater number of engines in the
500-750 kW and 1000-1500 kW size ranges than PSR estimates, whereas PSR’s
methodology predicts more than twice the number of permitted engines in the larger
1500+ kW size category. As one would expect given current state permitting thresholds,
PSR also estimates a larger number of small engines (<100 kW) than are indicated by
state permit records.
Table III-10
Maine Engine Data Comparison by Number and Capacity
PSR
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Total above permit size
29
120
175
117
20
25
9
65
560
119
TOTAL ENGINES
% State/PSR
State
Agreement
2
7%
78
65%
184
105%
158
135%
64
320%
28
112%
36
400%
28
43%
578
103%
156
131%
TOTAL CAPACITY (MW)
% State/PSR
PSR
State
Agreement
<1
<1
16%
10
6
58%
30
32
108%
43
56
130%
12
40
320%
20
23
119%
11
45
411%
129
60
47%
254
262
103%
172
168
98%
4. Massachusetts Inventory
Table III-11 summarizes PSR’s engine population and size distribution estimates for the
state of Massachusetts. According to these estimates, emergency engines account for
79% of all diesel engines and 75% of total engine capacity in Massachusetts.
Table III-11
PSR Estimates of Diesel Engines in Massachusetts by Number and Capacity
Number Totals Emergency Peak Baseload Total
25-50 kW
245
0
0
245
50-100 kW
837
207
13
1,057
100-250 kW
1,368
355
7
1,730
914
195
0
250-500 kW
1,109
500-750 kW
198
44
5
247
750-1000 kW
175
41
5
221
1000-1500 kW
101
121
4
226
156
20
16
1500+ kW
192
Total
3,994
983
50
5,027
Capacity Totals (MW) Emergency Peak Baseload
25-50 kW
8
0
0
50-100 kW
67
17
1
100-250 kW
234
59
1
328
70
0
250-500 kW
500-750 kW
123
28
3
750-1000 kW
139
32
4
1000-1500 kW
120
151
5
301
47
32
1500+ kW
Total
1,319
403
46
Total
8
85
295
398
153
175
275
379
1,768
The Massachusetts Department of Environmental Protection (MA DEP) provided
NESCAUM with two data sets of permitted engines. The first is a permit list that contains
information on all engines and facilities with fuel input greater than or equal to 10
MMBtu/hr. The earliest issued permits in the database are from 1998; thus it may fail to
include many older engines. MA DEP also provided a second list that included only
21
emergency engines. The emergency engine database did not specify the size of every
listed engine, though size could be inferred from other information for about half the
engines on the list. The resulting size distribution was assumed to be representative of
those engines whose size could not be determined. The left-hand side of Table III-12
summarizes state permit information for electricity generating IC engines in
Massachusetts; engines used to produce mechanical power for fire pumps and air
compressors are included on the right-hand side of Table III-12.
Table III-12
Massachusetts Permit Records for Electricity Generators and Mechanical Engines
Electricity
Generators
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Number of
Engines
11
13
278
156
138
73
160
275
1,104
Capacity
Totals (MW)
<1
1
38
54
77
61
188
602
1,021
Mechanical
Engines
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Number of
Engines
0
2
33
19
12
2
5
12
85
Capacity
Totals (MW)
0
0
5
6
7
2
6
35
59
Table III-13 compares PSR’s population estimates with data from Massachusetts permit
records. According to PSR’s estimation methodology, engines that fall below the
Massachusetts permitting threshold of 300 kW are likely to account for nearly 70% of all
engines in the state. Overall, state permit records would appear to account for just 22% of
PSR’s estimated engine population between 300 kW and 1500 kW for Massachusetts.
However, this population likely includes a number of emergency engines below 1 MW in
size that are eligible to operate under permit-by-rule, in which case they are not included
in the state database.
Table III-13
Massachusetts Engine Data Comparison by Number and Capacity
PSR
25-50 kW
50-100 kW
100-250 kW
250-300 kW
300-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Total above permit size
245
1,057
1,730
275
834
247
221
226
192
5,027
1,720
TOTAL ENGINES
% State/PSR
State
Agreement
11
4%
13
1%
278
16%
16
6%
140
17%
138
56%
73
33%
160
71%
275
143%
1,104
22%
786
46%
22
TOTAL CAPACITY (MW)
% State/PSR
PSR
State
Agreement
8
<1
5%
85
1
1%
295
38
13%
74
4
6%
324
50
15%
153
77
50%
175
61
35%
275
188
68%
379
602
159%
1,768
1,021
58%
1,307
978
75%
5. New Hampshire Inventory
Table III-14 summarizes PSR’s engine population and size distribution estimates for the
state of New Hampshire. According to these estimates, emergency engines account for
88% of all diesel engines and 85% of total engine capacity in New Hampshire.
Table III-14
PSR Estimates of Diesel Engines in New Hampshire by Number and Capacity
Number Totals Emergency Peak Baseload Total
25-50 kW
17
0
0
17
50-100 kW
130
15
0
145
100-250 kW
220
28
0
248
145
15
0
250-500 kW
160
500-750 kW
29
4
0
33
750-1000 kW
27
4
0
31
1000-1500 kW
6
7
0
13
81
8
7
1500+ kW
96
Total
655
81
7
743
Capacity Totals (MW) Emergency Peak Baseload
25-50 kW
1
0
0
50-100 kW
11
1
0
100-250 kW
38
5
0
53
6
0
250-500 kW
500-750 kW
18
2
0
750-1000 kW
21
3
0
1000-1500 kW
8
9
0
156
18
11
1500+ kW
Total
304
44
11
Total
1
12
42
59
20
24
16
185
359
The New Hampshire Department of Environmental Services (NH DES) provided
NESCAUM with a list of both emergency and non-emergency engines. Engine
characteristics – such as type, size and fuel – were specified in this list, but not engine
application. The left-hand side of Table III-15 summarizes information from New
Hampshire’s list of electricity-generating IC engines, while the right-hand side of Table
III-15 summarizes state information on 22 mechanical IC engines.
Table III-15
New Hampshire Permit Records for Electricity Generators and Mechanical Engines
Electricity
Generators
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Number of
Engines
1
2
65
126
71
39
47
9
360
Capacity
Totals (MW)
<1
<1
12
44
41
32
56
30
215
Mechanical
Engines
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Number of
Engines
1
0
12
5
4
0
0
0
22
Capacity
Totals (MW)
<1
0
2
2
2
0
0
0
6
Table III-16 compares PSR’s population estimates with data from New Hampshire’s
permitted engine list. The comparison suggests that few engines under the 150 kW
threshold are included in state records. For engines over 150 kW generally, state records
would appear to capture a larger fraction of the estimated population. However, the
numeric discrepancy between known and estimated engines varies considerably across
different size categories. In the 150-500 kW size range, PSR estimates a significantly
larger number of engines than is reflected in the state’s list; the same applies in the over
1500 kW size range, whereas the opposite is true in the 500 to 1500 kW size range.
23
Again, this result may be indicative of the difficulty of estimating engine populations
with the degree of specificity involved in these state-level analyses.
Table III-16
New Hampshire Engine Data Comparison by Number and Capacity
PSR
25-50 kW
50-100 kW
100-150 kW
150-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Total above permit size
17
145
67
181
160
33
31
13
96
743
514
TOTAL ENGINES
% State/PSR
State
Agreement
1
6%
2
1%
10
15%
55
30%
126
79%
71
215%
39
126%
47
362%
9
9%
360
48%
347
68%
TOTAL CAPACITY (MW)
% State/PSR
PSR
State
Agreement
1
<1
8%
12
<1
1%
8
1
15%
35
11
32%
59
44
75%
20
41
202%
24
32
134%
16
56
344%
185
30
16%
359
215
60%
339
214
63%
6. New Jersey Inventory
Table III-17 summarizes PSR’s engine population and size distribution estimates for the
state of New Jersey. According to these estimates, emergency engines account for 79% of
all diesel engines and 73% of total engine capacity in New Jersey.
Table III-17
PSR Estimates of Diesel Engines in New Jersey by Number and Capacity
Number Totals Emergency Peak Baseload Total
25-50 kW
453
0
0
453
50-100 kW
1,454
358
19
1,831
2,325
588
23
100-250 kW
2,936
250-500 kW
1,461
312
1
1,774
500-750 kW
326
83
12
421
750-1000 kW
290
78
15
383
1000-1500 kW
169
166
7
342
1500+ kW
189
86
0
275
Total
6,667
1,671
77
8,415
Capacity Totals (MW) Emergency Peak Baseload
25-50 kW
15
0
0
50-100 kW
115
30
2
393
97
3
100-250 kW
250-500 kW
525
112
0
500-750 kW
202
51
8
750-1000 kW
230
61
12
1000-1500 kW
202
206
9
1500+ kW
356
180
0
Total
2,038
737
33
Total
15
147
494
637
260
303
417
535
2,808
The New Jersey Department of Environmental Protection (NJ DEP) provided
NESCAUM with two lists of permitted engines. The first is a list of 5,016 emergency
engines, which includes 552 mechanical engines used to power fire pumps, air
compressors and water pumps. Of the remaining 4,464 engines, information on engine
size was available for a subset of 1,220 engines, or 27% of the total list. In the summary
information presented below, the size distribution of the remaining engines was imputed
based on this known subset. The second list provided by NJ DEP contains 1,771 nonemergency engines, 412 of which were removed because they are mechanical or non-IC
24
engine turbines and boilers. Information on engine size was available for 637 (47%) of
the remaining 1,359 generator engines; as with the emergency engines, the size
distribution of this subset was assumed to be representative of the entire group. Table III18 summarizes information from the NJ DEP engine lists, incorporating the size
assumptions noted above. Given the large number of engines for which size had to be
imputed, a caution is in order concerning the accuracy of the size distribution indicated in
Table III-18.
Table III-18
New Jersey Permit Records for Electricity Generators and Mechanical Engines
Electricity
Generators
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Number of
Engines
4
120
1,432
1,247
927
837
698
558
5,823
Capacity
Totals (MW)
0
12
224
416
527
664
875
1,158
3,876
Mechanical
Engines
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Number of
Engines
0
32
314
374
97
0
20
34
871
Capacity
Totals (MW)
0
3
52
132
55
0
24
51
316
Because NJ DEP’s engine lists do not include size information for many permitted
engines, it is difficult to make a meaningful comparison between the PSR estimates and
state records. If the size distribution from known engines is applied to all other engines in
the state lists (as described above), the results suggest that PSR’s methodology
substantially underestimates the New Jersey engine population over 500 kW. However,
given the multiple assumptions involved in both sets of data, it is difficult to place much
confidence in this comparison.
Table III-19
New Jersey Engine Data Comparison by Number and Capacity
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Total above permit size
TOTAL ENGINES
% State/PSR
PSR
State
Agreement
453
4
1%
1,831
120
7%
2,936
1,432
49%
1,774
1,247
70%
421
927
220%
383
837
219%
342
698
204%
275
558
203%
8,415
5,823
69%
6,131
5,699
93%
25
TOTAL CAPACITY (MW)
% State/PSR
PSR
State
Agreement
15
<1
1%
147
12
8%
494
224
45%
637
416
65%
260
527
202%
303
664
219%
417
875
210%
535
1,158
216%
2,808
3,876
138%
1,516
3,225
213%
7. New York Inventory
Table III-20 summarizes PSR’s engine population and size distribution estimates for the
state of New York. According to these estimates, emergency engines account for
approximately 79% of the total engine population and 74% of total capacity.
Table III-20
PSR Estimates of Diesel Engines in New York by Number and Capacity
Number Totals Emergency Peak Baseload Total
25-50 kW
827
0
0
827
50-100 kW
2,562
644
73
3,279
100-250 kW
4,060
1,025
62
5,147
2,575
578
6
250-500 kW
3,159
500-750 kW
569
157
29
755
750-1000 kW
505
137
28
670
1000-1500 kW
293
306
26
625
458
110
7
1500+ kW
575
Total
11,849
2,957
231
15,037
Capacity Totals (MW) Emergency Peak Baseload
25-50 kW
28
0
0
50-100 kW
203
53
6
100-250 kW
688
169
9
923
207
2
250-500 kW
500-750 kW
352
97
18
750-1000 kW
400
108
22
1000-1500 kW
352
377
32
877
242
13
1500+ kW
Total
3,822
1,254
103
Total
28
262
866
1,132
466
530
761
1,133
5,179
NESCAUM obtained two lists of engines permitted in New York. The first is a list of 311
permitted, non-emergency engines from the state DEC. It provides detailed information
on engine owner, location, model and size for 237 (76%) of the listed engines. The
second list is from the New York City DEP, and contains specific information on the
owner, location, model and size of 1,351 permitted emergency generator engines in New
York City. Because the New York City DEP database includes engines permitted over
the last 25 years, it should provide reliable information on emergency engines in the
NYC area. Information from both lists is combined and summarized in Table III-21.
Table III-21
New York Permit Records for Electricity Generators and Mechanical Engines
Electricity
Generators
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Number of
Engines
26
93
338
407
274
199
176
149
1,662
Capacity
Totals (MW)
1
7
57
141
158
163
206
268
998
Mechanical
Engines
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Number of
Engines
44
37
43
21
4
2
1
7
159
Capacity
Totals (MW)
2
3
7
7
2
2
1
13
36
Table III-22 compares PSR’s population estimates for New York with data from state and
city permit records. The total number of engines included in the state and city lists
described above account for only 11% of the engine population estimated by PSR for the
state of New York. The fact that information on emergency engines was only available
for New York City, and not for the state as a whole, combined with the fact that many
26
engines statewide are likely too small to trigger permitting requirements, probably
accounts in some measure for this discrepancy.
Table III-22
New York Engine Data Comparison by Number and Capacity
25-50 kW
50-100 kW
100-250 kW
250-300 kW
300-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Total above permit size
TOTAL DIESEL ENGINES
% State
PSR
State
Agreement
827
26
3%
3,279
93
3%
5,147
337
7%
1,072
82
8%
2,087
328
16%
755
272
36%
670
201
30%
625
175
28%
575
148
26%
15,037
1,662
11%
4,712
1,124
24%
TOTAL CAPACITY (MW)
% PSR
PSR
State
Agreement
28
1
2%
262
7
3%
866
57
7%
211
21
10%
922
119
13%
466
158
34%
530
163
31%
761
206
27%
1133
268
24%
5179
998
19%
3812
913
24%
8. Rhode Island Inventory
Table III-23 summarizes PSR’s engine population and size distribution estimates for the
state of Rhode Island. According to these estimates, emergency engines account for
approximately 91% of both total engine population and capacity.
Table III-23
PSR Estimates of Diesel Engines in Rhode Island by Number and Capacity
Number Totals Emergency Peak Baseload Total
25-50 kW
15
0
0
15
50-100 kW
81
8
0
89
100-250 kW
111
13
0
124
79
5
0
250-500 kW
84
500-750 kW
14
0
0
14
750-1000 kW
17
1
0
18
1000-1500 kW
3
4
0
7
12
0
0
1500+ kW
12
Total
332
31
0
363
Capacity Totals (MW) Emergency Peak Baseload
25-50 kW
0
0
0
50-100 kW
7
1
0
100-250 kW
19
2
0
29
2
0
250-500 kW
500-750 kW
9
0
0
750-1000 kW
13
1
0
1000-1500 kW
4
5
0
25
0
0
1500+ kW
Total
106
11
0
The data set available from Rhode Island’s Department of Environmental Management
(RI DEM) contains 86 engines, 14 of which are used to produce mechanical power
instead of electricity. Size information is available for all engines in the data set and is
summarized in Table III-24.
27
Total
0
7
21
31
9
14
9
25
117
Table III-24
Rhode Island Permit Records for Electricity Generators and Mechanical Engines
Electricity
Generators
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Number of
Engines
0
0
4
1
20
11
11
25
72
Capacity
Totals (MW)
0
0
1
<1
12
10
13
139
174
Mechanical
Engines
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Number of
Engines
0
0
0
0
11
0
0
3
14
Capacity
Totals (MW)
0
0
0
0
7
0
0
6
13
Table III-25 compares PSR’s estimates to the information available from Rhode Island’s
permit records. Not surprisingly, given current permitting thresholds, few of the smaller
engines estimated by PSR have been permitted by the state. For larger engines, PSR
appears to have underestimated the actual Rhode Island engine population in all but the
750-1000 kW size bracket.
Table III-25
Rhode Island Engine Data Comparison by Number and Capacity
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Total above permit size
TOTAL ENGINES
% PSR
PSR
State
Agreement
15
0
0%
89
0
0%
124
4
3%
84
1
1%
14
20
143%
18
11
61%
7
11
157%
12
25
208%
363
72
20%
51
67
131%
TOTAL CAPACITY (MW)
% PSR
PSR
State
Agreement
<1
0
0%
7
0
0%
21
1
3%
31
<1
1%
9
12
132%
14
10
68%
9
13
150%
25
139
554%
116
174
149%
57
173
306%
9. Vermont Inventory
Table III-26 summarizes PSR’s engine population and size distribution estimates for the
state of Vermont. According to these estimates, emergency engines account for
approximately 93% of engines and 89% of total engine capacity in Vermont.
28
Table III-26
PSR Estimates of Diesel Engines in Vermont by Number and Capacity
Number Totals Emergency Peak Baseload Total
25-50 kW
14
0
0
14
50-100 kW
72
7
0
79
100-250 kW
100
9
0
109
70
2
0
250-500 kW
72
500-750 kW
13
0
0
13
750-1000 kW
16
1
0
17
1000-1500 kW
2
4
0
6
0
0
0
1500+ kW
0
Total
287
23
0
310
Capacity Totals (MW) Emergency Peak Baseload
25-50 kW
0
0
0
50-100 kW
6
1
0
100-250 kW
17
2
0
26
1
0
250-500 kW
500-750 kW
8
0
0
750-1000 kW
13
1
0
1000-1500 kW
3
5
0
0
0
0
1500+ kW
Total
73
9
0
Total
0
7
18
27
8
13
8
0
81
For this report, NESCAUM reviewed information from the Vermont Department of
Environmental Conservation (VT DEC) including Title V permits, state operating
permits, pre-construction permits and file correspondence. Review of the Title V permits
yielded detailed information regarding the location of generators, their size and fuel type.
Of the 216 engines described in state records, 66 are used to generate electricity and 150
produce mechanical power. The latter category includes large IC engines used at lumber
mills, ski resorts and quarries. Details on both engine populations are summarized in
Table III-27 below.
Table III-27
Vermont Permit Records for Electricity Generators and Mechanical Engines
Electricity
Generators
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Number of
Engines
0
9
18
17
7
2
10
3
66
Capacity
Totals (MW)
0
1
3
5
4
2
11
6
32
Mechanical
Engines
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Number of
Engines
1
14
31
73
9
14
8
0
150
Capacity
Totals (MW)
<1
1
6
24
5
13
9
0
58
Table III-28 compares information available from the VT DEC with PSR’s population
estimates. The comparison suggests that the state may have permitted only about a third
of the total engine population estimated by PSR to meet current permitting thresholds.
29
Table III-28
Vermont Engine Data Comparison by Number and Capacity
25-50 kW
50-100 kW
100-250 kW
250-337 kW
337-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Total above permit size
TOTAL ENGINES
% PSR
PSR
State
Agreement
14
0
0%
79
9
11%
109
18
17%
24
12
50%
48
5
10%
13
7
54%
17
2
12%
6
10
167%
0
3
--310
66
21%
84
27
32%
30
TOTAL CAPACITY
% PSR
PSR
State
Agreement
<1
0
0%
7
1
11%
18
3
17%
7
4
50%
20
2
9%
8
4
47%
13
2
12%
8
11
146%
0
6
--81
32
39%
49
24
50%
IV. Engine Population Surveys – New York City and
Fairfield County, CT
A. Introduction
As a follow-up and complement to the inventory efforts described in Chapter III, Power
Systems Research (PSR) conducted detailed telephone surveys to obtain information on
distributed generator populations in New York City and Fairfield County, Connecticut.
These areas were selected because both face severe transmission constraints and have
been the focus of recent efforts by electric system operators to encourage customer-side
demand responses during periods of peak electrical demand. As indicated in Chapter II,
customer-side demand response can include efforts to curtail load (e.g. turning off
equipment), increased on-site generation, or a combination of both. Thus, both New York
City and Fairfield County are areas where the potential for increased use of diesel engines
as a supplemental power source during periods of peak demand is likely to be higher than
in other areas of the Northeast. For Fairfield County, in particular, this potential may be
further increased by the recent introduction of locational marginal pricing in the New
England power pool. Because locational marginal pricing is designed to more closely
reflect the marginal price of supplying grid power to specific areas within the larger
power pool, it will tend to enhance the price signals promoting demand response in
transmission-constrained areas.
After briefly reviewing the methodology used by PSR in conducting the telephone
surveys, this chapter reviews the survey results and compares them to permit data on file
with the relevant state environmental agencies.
B. Methodology
The first step in conducting the telephone survey was to identify likely owners of diesel
generators. To do so, PSR applied the same basic estimation methodology used to
develop the state-level population estimates described in the previous chapter.19 Using
business data from Fairfield County and New York City, PSR developed an initial
estimate of the likely number of units in each area and identified potential owners to be
contacted.20 This information was used to help prioritize survey calls in terms of the
highest incidence group and successively lower probability owners. In general, PSR
started with phone calls to businesses having more than 50 employees and with SIC
codes identified as higher probability for owning on-site generators. If this initial sample
was exhausted before the survey quota was reached, additional calls were made to
19
The description of PSR’s Partslink methodology in Appendix A, includes a section (starting on page 3)
entitled “Application to the NESCAUM Survey” that describes the survey methodology in more detail.
20
In some cases, owners identified for this effort had previously been surveyed by PSR in the development
of its Partslink database.
31
businesses with less than 50 employees until PSR identified more than 70-75% of the
number of engines estimated to be in the area.
PSR initially calculated that the likely generator populations for the five-county New
York City area and for Fairfield County totaled more than 24,000 and 5,000 units,
respectively. However, these initial estimates included both portable and stationary
generator sets. Of the total generator population estimated for New York City, PSR’s
methodology indicated that more than 18,000 units were likely to be smaller than 10 kW
in size; the comparable estimate for very small units in Fairfield County was 4,300. All of
these smaller than 10 kW units and half of the remaining estimated generator population
up to a capacity of 300 kW were assumed to be portable and were excluded from the
survey sample.
PSR began contacting potential generator owners in the spring of 2002. In each call, PSR
sought to determine first whether an engine was present and, if so, to follow-up with a
series of questions about engine application and use, as well as size, age, fuel and hours
of operation during a typical 12-month period.21 Survey efforts were eventually
completed in August 2002, at which point PSR had identified more than 1,700 generator
sets owned by 1,536 organizations in New York City and 292 generator sets owned by
227 organizations in Fairfield County.
C. Telephone Survey Results
1. New York City
PSR contacted 2,475 businesses in New York City, of which 1,536 were found to own a
total of 1,724 generators. According to surveyed owners, 1,075 (62%) of these engines
were installed for emergency use. Current state permitting requirements specify that
emergency engines may only be used in the case of actual or imminent electric supply
shortfalls and are limited to less than 500 hours of operation each year. The great
majority of engines surveyed in New York City (1,440 engines or 84% of the total
sample of 1,724 units) run on diesel fuel; of these diesel engines, 903 (63%) are for
emergency use only. Tables IV-1 and IV-2 summarize the distribution of engine sizes,
fuel sources and potential applications for engines identified in the New York City
telephone survey.
21
Due to security concerns, PSR encountered significant difficulties collecting information – especially
from New York City businesses – after September 11, 2001. To obtain information, it became necessary in
some cases to send formal letters to target businesses explaining the purpose of the survey and urging
cooperation.
32
Table IV-1
New York City PSR Survey Results by Engine Fuel and Application
Diesel
Natural Gas
Gasoline
Total
Emergency
903
149
23
1,075
Peak
10
2
0
12
Baseload
527
98
12
637
Total
1,440
249
35
1,724
Table IV-2
New York City PSR Survey Results by Engine Size and Capacity
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Engine Totals
5
82
257
379
334
322
319
26
1,724
% Engines
0.3%
4.8%
14.9%
22.0%
19.4%
18.7%
18.5%
1.5%
100.0%
Total Capacity (MW)
0
4
39
135
202
272
346
56
1,054
Table IV-2 indicates that more than half (58%) of the engines identified in PSR’s survey
are larger than 500 kW, most operate on diesel fuel and most are intended for emergency
use. The total generating capacity of all engines identified in the New York City
telephone survey is 1,054 MW. This compares to a peak summer demand for the New
York City area of approximately 11,000 MW in 2003.22
In addition to information on numbers and sizes of engines in place, the telephone survey
was used to collect information on engine operation in a typical 12-month period.
Reported hours of operation were then combined with information on engine capacity to
calculate electrical generation in MWh per year. The results of this calculation, by engine
fuel and application, are summarized in Table IV-3 below. Overall, the 1,724 engines
identified in the survey were estimated to generate a total of 490,000 MWh in a typical
year.23 Non-emergency operation (baseload and peak operation) accounted for 80% of
this generation total, while diesel engines accounted for 72% (355,000 MWh) of total
reported generation.
Table IV-3
New York City PSR Survey Generation Totals in MWh/yr by Fuel and Application
22
In February 2003, the New York ISO forecast peak summer demand for 2003 at 11,020 MW
(http://www.nyiso.com/topics/articles/news_releases/2003/nr_022503_summer_outlook.pdf).
23
The 490,000 MWh total is the equivalent of a 100 MW power plant running 60% of the time, or 5000
hours/yr.
33
Diesel
Natural Gas
Gasoline
Total
Emergency
84,401
14,830
1,749
100,981
Peak
2,524
291
0
2,815
Baseload
268,258
111,690
5,614
385,562
Total
355,183
126,811
7,363
489,357
2. Fairfield County, CT
PSR contacted 1,557 businesses in Fairfield County, Connecticut, of which 227 were
found to own a total of 294 generators. Nearly all of these engines (280) are designated
for emergency use only (meaning they can operate only in cases of actual or imminent
outages, when the ISO has officially called for voltage reductions) and most (195) are
diesel-powered. Of the 14 non-emergency engines identified, none were diesel powered.
Rather, natural gas fueled units accounted for most of the generators designated for peak
shaving and baseload purposes, with gasoline powered engines making up the remaining
non-emergency population.
Table IV-4
Fairfield County PSR Survey Results by Engine Fuel and Application
Diesel
Natural Gas
Gasoline
Total
Emergency
195
65
20
280
Peak
0
5
1
6
Baseload
0
6
2
8
Total
195
76
23
294
Table IV-5
Fairfield County PSR Survey Results by Engine Size and Capacity
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Engine Totals % Engines
18
6.1%
72
24.5%
79
26.9%
86
29.3%
21
7.1%
4
1.4%
7
2.4%
7
2.4%
294
100%
Capacity Totals (MW)
1
5
12
27
11
3
8
22
88
In both numbers and total capacity, the generator population in Fairfield County was
smaller than that in New York City, as seen in Tables IV-4 and IV-5. In addition, the
Connecticut units were generally smaller, with 85% reporting capacity ratings less than
500 kW.
As in New York City, the telephone surveys were used to collect additional information
on hours of operation for surveyed engines. The resulting estimates of electrical
production by these engines are summarized in Table IV-6. They indicate that the 294
34
engines identified in the survey generate a total of 9,300 MWh in a typical 12-month
period.24 More than half (62%) of this total output for a typical year was generated by the
8 baseload and 6 peak shaving engines. The remaining 3,500 MWh was generated by
emergency engines.
Table IV-6
Fairfield County PSR Survey Generation Totals in MWh/yr by Fuel and Application
Diesel
Natural Gas
Gasoline
Total
Emergency
2,860
563
89
3,512
Peak
0
834
100
934
Baseload
0
2,249
2,628
4,877
Total
2,860
3,646
2,817
9,323
D. Overlap between Survey Results and State Permitting Records
To better assess the extent to which current state permitting requirements and records
capture the population of surveyed engines, the results of PSR’s telephone surveys were
compared to permitting records compiled by the New York City and State of Connecticut
Departments of Environmental Protection (DEPs). The results of this comparison for the
two survey areas are discussed below.
1. New York City
Although engines in New York City are regulated by both the State Department of
Environmental Conservation (DEC) and the New York City DEP (see discussion in
Chapter II), the permit information NESCAUM used to compare survey results for New
York City are available only from the City DEP. The City DEP has permit information on
1,351 emergency engines. As indicated in Chapter II, New York City DEP regulations
require that diesel units down to a size threshold of 33 kW (350,000 Btu/hr) be registered
with the City. Engines over 280 kW (2.8 MMBtu/hr) must obtain a work permit unless
they are emergency engines, in which case they must simply register. The state requires
permits for all engines over 160 kW (225 hp) in the New York City metropolitan area.
Emergency engines are limited to a total of 500 hours of operation per year (of which 200
hours may be used when called by the New York ISO as part of its emergency demand
response program; that is, when outage is imminent, as opposed to when it has already
occurred).25
Table IV-7 summarizes the size distribution of emergency engines registered with the
New York City DEP. More than half the engines are smaller than 500 kW; median size is
413 kW.
24
The 9,300 MWh total is equivalent to a 2 MW power plant running 60% of the time, or for 5000 hrs/yr.
Because the NY ISO’s emergency demand response program is limited to situations of imminent
shortfall, it may be distinguished from peak shaving or price-responsive demand response programs.
Emergency generators in NYC are prohibited from operation for peak shaving or price response reasons.
25
35
Table IV-7
New York City DEP Permitted Emergency Engines
25-50 kW
50-100 kW
100-250 kW
250-300 kW
300-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Engine Totals
26
90
313
69
244
195
148
142
124
1,351
% Engines
1.9%
6.7%
23.2%
5.1%
18.1%
14.4%
11.0%
10.5%
9.2%
100%
Capacity Totals (MW)
1
6
52
18
89
111
119
168
222
786
Comparing the specific New York City engines identified in PSR’s telephone survey to
the DEP database, PSR found that 839 engines (based on owner, address and size) were
included in both lists. This suggests that 512 engines registered with the City were not
captured by the telephone survey. Conversely, the comparison indicates that 885 engines
identified in the survey are not registered with the City DEP, 796 of which should be
registered according to the regulations outlined above. Combining both databases, a total
of 2,236 individual engines were identified in New York City. The size distribution of
this combined population is summarized in Table IV-8 below.
Table IV-8
Total Engine Inventory for New York City
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Engine Totals
26
123
426
509
389
328
319
116
2,236
% Engines
1.2%
5.5%
19.1%
22.8%
17.4%
14.7%
14.3%
5.2%
100%
Capacity Totals (MW)
1
8
70
178
229
277
346
211
1,320
2. Fairfield County, Connecticut
The state of Connecticut maintains extensive permitting data on distributed generator
engines. As described in more detail in Chapter II, current state permit requirements
distinguish between engines intended for emergency use only and other engines. In
addition, lower permitting thresholds have recently been instituted for the transmissionconstrained southwestern Connecticut load pocket (including Fairfield County). In this
area, any emergency engine over 37 kW (50 hp) is required to obtain a permit. Thus,
Connecticut’s permit records should capture most engines over 500 kW throughout the
state, and many more of the smaller engines in southwestern Connecticut. However, the
36
permit requirements that established the 37 kW threshold were new in 2002, and only 10
additional engines were permitted under them as of March 2003. As indicated by Table
IV-9 below, state permitting records imply that 58% of permitted engines are smaller
than 500 kW (as opposed to 85% in the PSR survey data).
Table IV-9
Connecticut DEP Permitted Engine Summary for Fairfield County
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Engine Totals
11
40
58
75
54
22
29
26
315
% Engines
3.5%
12.7%
18.4%
23.8%
17.1%
7.0%
9.2%
8.3%
100%
Capacity Totals (MW)
0
3
9
26
31
18
33
63
184
PSR compared the results of its Fairfield County telephone survey with state permit
records for engines located in Fairfield County. The comparison found an overlap of only
49 individual engines between the two databases. Combining all individual engines
identified in either the telephone survey or state permit records results in a total database
of 560 engines with a combined capacity of 235 MW (see Table IV-10).
Table IV-10
Total Engine Inventory for Fairfield County
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Engine Totals % Engines
29
5.2%
108
19.3%
135
24.1%
148
26.4%
66
11.8%
19
3.4%
29
5.2%
26
4.6%
560
100%
Capacity Totals (MW)
1
7
21
48
37
16
32
74
235
3. Analysis
The fact that there is so little overlap between the PSR telephone survey data and permit
records for New York City and Fairfield County suggests that neither the survey
methodology used by PSR nor current state permitting systems can be relied upon for
comprehensive coverage of the existing population of small generators. The percentage
of surveyed engines for which there was no record in state or city databases was over
50% in the case of New York City and over 80% in the case of Fairfield County. While
the percentage overlap is smaller for Fairfield County, the number of unregistered
37
engines identified in the telephone survey was actually larger in New York City (885
units compared to 245 in Fairfield County).
The fact that the telephone surveys did not capture all permitted engines is not
particularly surprising given a number of known flaws in the survey approach. In the first
place, the estimation methodology used to identify likely generator owners was inexact.
Moreover, the list of target businesses that PSR began with seems to have excluded
certain types of institutions, such as government agencies and healthcare facilities, which
are likely to have on-site generators. (The list used by PSR came from American
Business Lists, and after comparison with the permitted engines it was evident that a few
SIC categories were not listed on the primary business list. The reason for this is not
known, but PSR has concluded that its results nevertheless represent statistically valid
estimates.26) Additional inaccuracies were introduced by the fact that PSR targeted likely
generator owners, rather than likely facilities or locations. In some cases a single owner
might own multiple generators at different locations and might not have identified them
all when contacted. Or the owner of a generator installed within the survey area might not
have been contacted because the owner was listed as having a different location outside
the survey area. Finally, some contacted businesses did not respond, while others claimed
not to have a generator engine on site despite permit information to the contrary.
The fact that the telephone survey identified significant numbers of engines that were not
included in state databases is therefore of more concern than the fact that the survey
appears to have missed a number of engines on record with state agencies. Unfortunately,
it is difficult to determine precisely which engines located in the telephone survey should
be – but aren’t – either permitted or registered with the relevant state or city authorities.
The list of engines available from the New York City DEP, for example, included only
permitted emergency engines. As such, it did not include any engines that might qualify
for a permit-by-rule or any non-emergency engines. A list of non-emergency engines was
not available from the City DEP, while the State DEC’s database on permitted nonemergency engines did not provide sufficient information to determine which engines
were located within New York City. In Fairfield County, many of the 245 engines that
were identified in the survey but that are not included in the state’s database fall below
the CT DEP’s permitting threshold of 500 kW. Nearly all of these engines would be
subject to the lower 37 kW permitting threshold for engines in southwestern Connecticut.
However, as noted previously, the regulations introducing this lower permitting threshold
did not take effect until 2002, and had resulted in the permitting of only 10 additional
engines in southwestern Connecticut as of March 2003.
26
Specifically, the description of PSR’s methodology in Appendix A states: “As a result, our compilation
is probably over representative in some SIC’s and under representative in others. Nonetheless, the results
and incidence of ownership in the target areas were similar enough to the national pattern to make
statistically valid estimates of the installed generator set population.”
38
V.
Emissions Estimates
This study was motivated by a lack of information on the existing base of distributed
generation capacity in the Northeast and by a corollary lack of information on the
potential environmental impacts associated with increased use of that capacity under
changing electricity market conditions. As foregoing chapters have indicated, the existing
base of distributed generation capacity in the Northeast and elsewhere in the nation is
overwhelmingly dominated by diesel internal combustion engines, which typically have
very high emissions rates for several pollutants of concern. All or nearly all of these
diesel engines are also most likely to operate during periods of peak electricity demand,
whether in response to an emergency shortfall in grid-supplied electricity, to assist a
utility in avoiding such a shortfall, or simply to avoid the high prices that may occur
during peak periods. Throughout most of the Northeast, peak electricity demand occurs in
the summer time, during the hottest hours of the day. These are also the times when air
quality is likely to be at its worst and states and localities are most apt to be in violation
of federal ambient air quality standards. Because small diesel generators are often located
near or in densely populated urban areas, and because their emissions – which include
toxic constituents and high levels of particulate matter, as well as pollutants like nitrogen
oxides (NOx), hydrocarbons or volatile organic compounds (HC or VOC),27 sulfur
dioxide (SO2) and carbon monoxide (CO) – tend to be released closer to the ground,
operation of these engines, especially during peak hours, poses particular public health
concerns. Because no mechanism has existed to comprehensively track these units,
empirical data on their actual historic emissions or potential future impact have been
scarce. Hence, an important aspect of the survey effort was the collection of information
on engine operation.
This chapter begins by providing some additional detail on the particular health risks
associated with diesel exhaust. Although many previous analyses and many existing state
regulatory provisions focus on NOx emissions,28 other components of the exhaust
emissions from a diesel generator are likely to pose more significant air quality and
public health risks, particularly in the immediate vicinity of such a generator. In fact,
emissions of particulate matter and toxic or hazardous air pollutants are likely to be
27
The term “hydrocarbons” encompasses a very large number of chemical compounds characterized by
different molecular combinations of hydrogen and carbon. Together, nitrogen oxides (NOx) and
hydrocarbons are the primary precursor pollutants in the photochemical formation of ground-level ozone
(commonly known as “smog”). Technically, VOCs are a subset of hydrocarbons. Additional terms that are
used for this class of pollutants in some regulatory settings include NMHC (non-methane hydrocarbons),
NMOG (non-methane organic gases) and NMOC (non-methane organic compounds). While the specific set
of chemical compounds encompassed by each of these terms and acronyms differs slightly, we utilize the
terms VOC and HC more or less interchangeably throughout this report.
28
The focus on NOx is likely due to the fact that most of the Northeast’s non-attainment problems with
respect to federal health-based air quality standards have centered on ground-level ozone. Nitrogen oxides,
together with hydrocarbons, are the chief pollutants responsible for ozone pollution. As noted in the
introduction to this report, NOx emissions for uncontrolled diesel generators are far higher on a per MWh
basis than for most other conventional generation options. However, total NOx emissions from distributed
diesel electricity generators to date are probably relatively small compared to other NOx sources in most
state inventories. Hence the discussion in this chapter focuses on diesel particulate emissions.
39
responsible for some of the most significant health impacts associated with diesel IC
generators.
The second section of this chapter describes the results of a preliminary emissions impact
analysis using data collected in the Power Systems Research (PSR) telephone survey, as
well as initial assessments of the emissions impacts associated with on-site generators
operated as part of recent summer demand response programs sponsored by the New
York and New England ISOs. Overall, these results indicate that emissions from on-site
generators operated under these programs to date are relatively small compared to overall
emissions inventories for pollutants such as NOx and PM10 at the state or regional level.29
For comparison with the emissions estimates presented later in this chapter, for example,
the U.S. Environmental Protection Agency’s (EPA’s) estimated 1999 NOx and PM10
inventories for the state of New York totaled over 830,000 tons and 558,000 tons,
respectively. The estimated 1999 emissions inventory for these pollutants in Connecticut
was over 140,000 tons for NOx and almost 64,000 tons for PM10.30
However, any comparisons based on annual or seasonal emissions totals fail to capture
important dimensions of the public health concern, which include the spatial and
temporal concentration of distributed generator emissions as well as the toxic and
potentially carcinogenic nature of diesel emissions in particular. Another study, currently
underway with funding from EPA, is attempting to model the emissions impacts of
demand response programs in New England and has come to similar preliminary
conclusions: that net emissions from on-site generators operating under formal demand
response programs are likely to be small in most years compared to total electric system
emissions, but that these emissions could nevertheless pose local health threats in cases
where generators are located in densely populated areas.31
A. Health Risks Associated with Diesel Exhaust
Diesel exhaust is a complex mixture of gases and particles. Its characteristics – in terms
of chemical composition and constituent particle sizes – vary significantly across
different engine types, engine operating conditions and fuel formulations. The gaseous
fraction of diesel exhaust is composed of typical combustion gases as well as hazardous
air pollutants including volatile organics, alkenes, aromatic hydrocarbons, and
aldehydes.32 One of the main characteristics of diesel engines is the release of tiny
29
PM10 refers to particulate matter 10 microns in aerodynamic diameter or smaller. Elsewhere in this report
we refer to “fine” particles or PM2.5, which includes particles 2.5 microns in aerodynamic diameter or
smaller.
30
This emissions information is from the EPA’s 1999 National Emissions Inventory.
31
The analysis: Estimating Emissions from Demand Response Programs in New England, is being
conducted by Synapse Energy Economics with EPA funding. Its results are expected to be released during
the summer of 2003.
32
EPA has collected and reviewed emissions data on hazardous air pollutants (HAPs) as part of its
proposed MACT rulemaking on reciprocating internal combustion engines (RICE), available at
http://www.epa.gov/ttn/atw/mactprop.html. Additional information on HAP emissions from IC engines is
available from EPA’s Industrial Combustion Coordinated Rulemaking (ICCR) Federal Advisory
40
particles that are mainly aggregates of spherical carbon particles coated with inorganic
and organic substances. These substances consist of elemental carbon and soluble
compounds such as aldehydes, alkanes and alkenes, and high-molecular weight
polycyclic aromatic hydrocarbons (PAH) and PAH-derivatives.
A large percentage of the U.S. population is exposed to ambient fine particles (PM2.5), 33
of which diesel particulate matter is typically a significant constituent. An extensive
epidemiological literature – including several major studies – provides evidence of the
adverse health effects of airborne particles on human health. Short and long-term
exposures to fine particles are associated with acute and chronic excess morbidity and
mortality in the general population. Vulnerable subgroups include those individuals who
have existing respiratory or lung inflammation, repeated respiratory infections, or chronic
bronchitis or asthma. Children and the elderly may also have increased sensitivity to
PM2.5 exposure. Recent epidemiological research suggests that sub-daily exposures on a
time-scale of 1-5 hours can produce heart attacks and exacerbate allergenic responses.
Available evidence indicates that there are significant human health hazards associated
with exposure to diesel exhaust.34 EPA recently concluded that diesel exhaust is a chronic
respiratory hazard and a probable human lung carcinogen.35 Documented short-term
responses to exposure to diesel exhaust include pulmonary resistance (i.e. difficulty
breathing due to airway constriction), acute irritation (e.g. eye, throat, bronchial),
neurophysiological symptoms (e.g. lightheadedness, nausea and slowed reaction time, as
well as difficulty with balance, verbal recall and color perception), and respiratory
symptoms (cough, phlegm). Epidemiological studies indicate that occupational exposure
to diesel exhaust can result in increased frequency of bronchitic symptoms, cough and
phlegm, wheezing, and decrement in lung function. Other studies have documented
health effects in children exposed to diesel particulates at school and in people who live
within 100 meters of highways that are heavily traveled by diesel trucks. Children with
bronchial hyper-reactivity or susceptibility to other allergens appear to be particularly
sensitive to adverse effects. These findings are supported by studies of laboratory animals
which have demonstrated that the inhalation or direct application of diesel into the
respiratory tract – over both chronic and short-term exposures – induces inflammatory
airway changes, lung function changes, and increased susceptibility of exposed animals
to lung infection.
Committee Act (FACA) process (at http://www.epa.gov/ttnatw01/iccr/engine/em97rice.pdf) and from a
study published by the Western States Petroleum Association and the American Petroleum Institute titled
Air Toxic Emission Factors for Combustion Sources Using Petroleum-Based Fuels (October 17, 1997).
33
See Footnote 29 for an explanation of the term “PM2.5”.
34
For further detail and additional references see, for example: Health Assessment Document for Diesel
Engine Exhaust. U.S. Environmental Protection Agency. Washington DC, 2002; Lloyd A.C. and Cackette
T.A. Diesel engines: environmental impact and control; Journal of the Air & Waste Management
Association [J. Air & Waste Manage. Assoc.] 2001, 51:809-847; or Staff Report: Initial Statement of
Reasons for Rulemaking—Proposed Identification of Diesel Exhaust as a Toxic Air Contaminant.
California Air Resources Board: Sacramento, CA, 1998.
35
The potential cancer-causing (carcinogenic) effects of exposure to diesel exhaust have been the subject
of numerous studies, involving both human subjects and in vitro laboratory investigations. These studies
have provided compelling qualitative evidence for the carcinogenic effects of diesel exhaust, though cancer
risks have proved difficult to quantify.
41
Finally, there are a number of review articles which postulate that air pollutants generally
– and diesel exhaust in particular – play a role in the increasing prevalence of asthma and
other allergic respiratory diseases. In studies with human volunteers, diesel exhaust
particles made people with allergies more susceptible to allergens such as dust and
pollen. Exposure to diesel exhaust also causes inflammation in the lungs, which may
aggravate chronic respiratory symptoms and increase the frequency or intensity of asthma
attacks. The potential relevance of these immunological endpoints to public health is very
high, due the high incidence of respiratory allergies and asthma in many urban areas.
B. Emissions Analysis for New York City and Fairfield County, CT
Using PSR Telephone Survey Results
This section describes an initial analysis of emissions impacts from distributed generators
in New York City and Fairfield County, Connecticut using the results of the PSR
telephone survey and city or state permitting records for these areas (see Chapter IV).
Because of the enormous uncertainties associated with the more general estimates of
engine population presented in Chapter III and because information on actual engine
operation is simply not available for the broader region, the analysis is confined to those
areas covered by PSR’s detailed telephone surveys. However, even these data contain
significant uncertainties with respect to total populations and operating hours. Given
these uncertainties, and the fact that the electricity markets in both New York City and
Fairfield County are likely to be somewhat unique – in terms of their transmission and
supply situation, certainly, and perhaps for other reasons as well – the applicability of
these results to other parts of the Northeast is uncertain.
1. Methodology for Emissions Impact Analysis
Engine owners contacted in the PSR telephone survey were asked to estimate actual
hours of operation per year for each of their units. Their responses were compiled to
estimate annual generation totals for each survey area (in MWh), aggregated by fuel type
and engine size. To reflect the fact that the telephone surveys captured only a subset of
the total engine population, engines identified in state or city databases – and not
identified in the surveys – were added to those identified through the telephone surveys
and were assumed to operate for a number of hours comparable to surveyed engines used
for the same application. The resulting adjusted generation totals were then combined
with emissions factors published by the Sacramento Metropolitan Air Quality
Management District (SMAQMD), which primarily cites EPA’s AP-42 values (see Table
V-1).36 The AP-42 emissions factors are from engine tests performed as many as 15 years
36
The emissions factors in Table V-1 are taken from the SMAQMD Internal Combustion Engine Policy
Manual, September 2002. The SMAQMD Manuel largely relies on EPA’s Compilation of Pollutant
Emission Factors, AP-42, Fifth Edition, Volume I: Stationary Point and Area Sources (located at
http://www.epa.gov/ttn/chief/ap42/index.html) except where emissions testing performed by the state of
California for purposes of BACT determinations revealed higher emissions rates than the AP-42 factors.
Additional sources of information on NOx emission factors and control costs include an Ozone Transport
Commission study conducted in support of its model rule for distributed generation (Control Measure
42
ago, and therefore do not necessarily reflect the performance of today’s engines. Given
that the average age of units identified in the telephone surveys was 12-14 years, the use
of emissions factors typical of new engines would likely understate actual emissions
impacts.
Compared to the emissions factors in Table V-1 for diesel engines larger than 600 hp, the
use of emissions factors typical of new diesel engines would reduce estimated NOx
emissions by approximately 30% and estimated PM10 by nearly 20%. Information on
emissions factors characteristic of new engines is available from the Regulatory
Assistance Project’s (RAP) Distributed Generation Emission Model Rule.37 While
emissions factors for new engines are readily available through emissions testing and
regulations, those for older engines more representative of the existing population are
more difficult to determine. The IC engine emissions factors used for this analysis are in
agreement with a recent Natural Resource Defense Council (NRDC) report, which cites a
range of emissions rates from 10 – 41 lb/MWh for NOx and 0.4 – 3 lb/MWh for PM10.38
Because uncontrolled diesel engines are at the high end of emissions rates for IC engines
generally, the NRDC figures are consistent with the emissions factors shown in Table V1. Additionally, historical emissions information from Caterpillar indicates that NOx
emissions rates for engines manufactured in the late 1980s and early 1990s averaged 32
lb/MWh.39 Finally, emissions data compiled by the state of New Hampshire from 45
stack tests on IC engines show NOx emissions rates varying from a low of 11.4 lb/MWh
to a high of 42.5 lb/MWh.
Table V-1
Distributed Generator NOx, PM10 and VOC Emissions Factors
Engine Type
Diesel < 600 hp
Diesel > 600 hp
Natural Gasa
Gasoline
a
b
NOx
g/hp-hr
lb/MWh
14.06
41.47
10.86
32.04
b
b
10.89
32.12
5.00
14.72
PM10
g/hp-hr
lb/MWh
1.00
2.95
0.32
0.94
b
b
0.15
0.45
0.33
0.97
VOC
g/hp-hr
lb/MWh
1.14
3.36
b
b
1.00
2.95
b
b
0.43
1.27
9.79
28.88
Emissions factors represent an average for three types of natural gas engines. Additional detail in Table VI-2.
SMAQMD emission factors used in place of AP-42, see Footnote 36.
Development Support Analysis of OTC Model Rules, 2001 available at:
http://www.sso.org/otc/Publications/pub2.htm) and NESCAUM’s December 2000 report: Status of NOx
Controls for Gas Turbines, Cement Kilns, Industrial Boilers and IC Engines.
37
This information – which was compiled by Joel Bluestein – can be accessed online at:
http://www.raponline.org/ProjDocs/DREmsRul/Collfile/DGEmissionsMay2001.PDF.
38
From the NRDC article: “Small and Clean is Beautiful: Exploring the Emissions from Distributed
Generation and Pollution Prevention Policies,” Nathanael Greene and Roel Hammerschlag, Electricity
Journal, June 2000.
39
This is the average for all diesel engines at the time of manufacture and does not account for the effect of
normal wear and tear.
43
2. Estimated Emissions Impact for New York City
Survey results for New York City indicate that the units identified by PSR produce a total
of approximately 489,400 MWh in a typical year (as shown in Table IV-3), based on
reported hours of operation and individual unit capacities. Owners of emergency engines
contacted through the survey indicated that their engines typically operated about 150
hours per year. Assuming the same hours of operation apply to those 512 units in the
City’s list of emergency engines that were not captured in the telephone survey, results in
an additional 42,300 MWh of estimated electrical output. Combining these generation
totals with the emissions factors shown in Table V-1, yields the emissions estimates
shown in Table V-2. It should be emphasized, however, that considerable uncertainties
are embedded in these estimates. For example, PSR’s survey identified a number of
engines that were also included in the City’s list of emergency engines (even though
these engines were not identified as emergency engines in the survey), that seemed to
operate, according to the owners’ survey responses, well beyond the limits theoretically
applicable to emergency engines. Specifically, reported hours of operation for the subset
of units common to both lists averaged approximately 450 hours per year, which – if
accurate – suggests either that the City’s list includes some non-emergency engines or
that some emergency engines are being operated beyond their permit limits. If one
assumed that all 512 un-surveyed units in the City list operated 450 hours per year (rather
than the 150 hours per year assumed for emergency engines in Table V-2), our estimate
of total generation would increase by more than 84,700 MWh, while associated annual
emissions would increase by 1,430 tons for NOx, by 50 tons for PM10 and by 100 tons for
VOC.
Table V-2
Estimated New York City Emissions from Surveyed and Permitted Engines
According to PSR Survey Results
NOx
(tons/yr)
379,620
6,480
144,720
2,320
7,360
50
531,700
8,850
Engines MWh/yr
Diesel
Natural Gas
Gasoline
Total
1,652
549
35
2,236
PM10
(tons/yr)
260
30
5
295
VOC
(tons/yr)
580
90
110
780
3. Estimated Emissions Impact for Fairfield County, Connecticut
Based on owner-reported hours of operation, the engines surveyed by PSR in Fairfield
County generate a total of approximately 9,300 MWh per year (as reported in Table IV6). Total generation is relatively modest because most of the engines identified in the
Fairfield County survey were designated for emergency use only and were estimated by
their owners to operate, on average, only 42 hours per year. As discussed in Chapter IV,
an additional 266 engines (beyond those captured in PSR’s telephone survey) are
identified in state permitting records. Of these, 184 are designated for emergency use and
were assumed to operate for the same average of 42 hours per year. The remaining 82
44
non-emergency engines identified in state permitting records were assumed to operate for
an average of 3,790 hours per year, consistent with the estimated hours of operation
reported by owners of non-emergency engines in PSR’s survey. The results of these
assumptions, applied to the combined population of engines identified in the telephone
survey and in state permitting records, are shown in Table V-3. The table indicates a total
of nearly 320,000 MWh per year of generation by distributed generators in Fairfield
County and estimated annual emissions impacts of over 5,000 tons for NOx, almost 200
tons for PM10 and more than 500 tons for VOCs. It should be emphasized that these
figures assume that the high level of utilization (nearly 3,800 hrs/yr) reported by owners
of non-emergency engines contacted in PSR’s survey apply equally to all non-emergency
engines included in state permit records for Fairfield County. As such, these estimates are
quite uncertain and may substantially overstate actual generation and emissions totals for
southwest Connecticut.
Table V-3
Estimated Fairfield County Emissions from Surveyed and Permitted Engines
According to PSR Survey Results
Diesel
Natural Gas
Gasoline
Total
Engines
MWh/yr
447
90
23
560
313,040
3,700
2,820
319,560
PM10
NOx
VOC
(tons/yr) (tons/yr) (tons/yr)
5,220
190
470
60
0.80
2.5
20
1.4
40
5,300
192
513
4. Estimated Summertime Emissions Impact
The above estimates of total electrical output and associated emissions for distributed
generators in New York City and Fairfield County are presented on an annual basis. For
reasons discussed elsewhere in this report, actual output from distributed generators, and
associated emissions, are unlikely to be evenly distributed throughout the year. In
particular, both output and emissions from engines operated for peak shaving purposes or
as part of emergency or price-based demand response programs are likely to be
concentrated during the summer months. These are also the months when emissions are
likely to be of greatest concern from an air quality and public health perspective. By
contrast, operation of baseload units, as well as any operation of emergency back-up
generators for routine maintenance and testing purposes is likely to be distributed more
evenly throughout the year. To estimate summer-only emissions in New York City and
Fairfield County we assumed that all operation by baseload units, plus up to 50 hours per
year of operation reported for emergency units, is distributed evenly throughout the year.
In addition, all operation by peak-shaving engines and any operation above 50 hours per
year by emergency engines are assumed to occur during the 5 month period from May to
September. The resulting estimated summertime (or “ozone season”) emissions for New
York City and Fairfield County are 53% and 42% of the total annual emissions estimates,
respectively. The proportion of summer emissions in Fairfield County is very close to
5/12. This reflects the dominance of non-emergency engines from the state’s permit files
in the estimated generation totals. The permits do not make a distinction between peak
45
and baseload engines, so to estimate summer emissions all non-emergency engines were
assumed to operate as baseload units – that is to say, consistently throughout the year.
C. Emissions Estimates for ISO-Sponsored Demand Response
Programs
As noted elsewhere in this report, the Independent System Operators (ISOs) for both New
York and New England have instituted formal demand response programs in the last few
years to help address system adequacy and reliability concerns during summer peak
demand periods. In general, these programs provide incentives for customers to
voluntarily reduce their demand for grid-supplied electricity during periods of high prices
and/or when reserve margins (the difference between available grid-connected generating
capacity and load) become critically low. These conditions can occur system-wide or in
particular transmission-constrained “load pockets” within the ISO region. An individual
customer’s demand response capability can include the ability to curtail load (e.g. by
turning off equipment) or the ability to self-generate using an on-site generator, or a
combination of both.
1. New York ISO Program
During the summer of 2002, the New York ISO invoked its price-responsive load
program on a total of four days. This program is activated when the ISO forecasts a dayahead operating reserve deficiency. Compensation for participating in the program is a
minimum of $500/MWh, but may be higher if the locational market price per marginal
MWh rises higher. For the summer of 2002, the New York ISO signed up 1,702
customers, who registered 1,480 MW of collective demand response capability in the
form of both on-site generation and load curtailment. These totals included an unknown
number of residential customers who registered a combined demand response capability
of 20 MW through a third party aggregator.
Data for the four price-responsive load program events in New York during the summer
of 2002 are summarized in Table V-4. All four events included New York City along
with other areas of the state, but only data for participation in New York City are shown.
The two events in April occurred on unusually hot days, when two major transmission
lines were out of service. Because these events occurred before the usual start of the
season, program coordinators relied on customers who had signed up the previous year.
For this reason, statewide participation in the April events was smaller than during the
later events in July and August, though the demand response in New York City was still
substantial. Because of the unexpected nature of the April event, there was little advance
notice to allow participating customers to plan load curtailment activities. Hence, it is
estimated that the demand response elicited on April 17 and 18 was mostly attributable to
on-site generation. For the later price-responsive events, NY-ISO’s record of the
contribution of on-site generation (as opposed to load curtailment) is shown in Table V-4.
46
Unfortunately, information about the specific engines that supplied on-site generation
during NY ISO’s 2002 price-responsive load program events is not available. Table V-4
provides emissions estimates calculated by applying emissions factors typical of a larger
than 600 hp diesel engine. These emissions factors were used because the average size of
load reduction that customers signed up for under this program was 870 kW. Cumulative
emissions for the four price response events using this assumption total almost 14 tons of
NOx and 0.4 tons of PM10. If emissions factors for diesel engines smaller than 600 hp
were used, estimated NOx emissions would increase by 37% and estimated PM10
emissions would increase by 68%. The average compensation paid to demand response
participants during these events was the ISO’s floor price of $500 per MWh (market
prices did not exceed this minimum during any of the 2002 events).
Table V-4
Emissions Estimates for 2002 New York ISO Price-Responsive Load Program in
New York City
Date
04/17/02
04/18/02
07/30/02
08/14/02
Total
MWh
Generated
154.6*
191.5*
267.7
251.3
865.1
NOx
(tons)
2.48
3.07
4.29
4.03
13.86
PM10
(tons)
0.07
0.09
0.13
0.12
0.41
VOC
(tons)
0.23
0.28
0.39
0.37
1.28
*No data on generation vs. curtailment were available for the April events. The
response is assumed to be from generation due to the unexpectedness of the events,
which occurred before the start of summer.
2. New England ISO Programs
The New England ISO sponsors two demand response programs: an emergency demand
response program and a price response program. As its name implies, the emergency
demand response program is designed to reduce system loads at times when demand
threatens to exceed supply and brown-outs or black-outs are imminent. This program is
triggered when the ISO begins calling for voltage reductions (technically, Step 12 of the
ISO’s Operating Procedure Number 4). By contrast, the ISO’s price response program is
designed to elicit voluntary demand response at any time when the power pool’s hourly
forecast “Energy Clearing Price” (ECP) rises above $100 per MWh. Compensation for
both programs is a minimum of $100/MWh and is a maximum of $100 above the forecast
ECP, depending on other factors such as transmission congestion impacts.
Results for Summer 2001
In a report released in August 2002, the New England ISO presented an assessment of
emissions impacts associated with its summer 2001 demand response programs.40 Table
40
The report, initially prepared by Boston-based EFI, a consulting firm, was released by ISO-NE in August
2002 under the title: Summer 2001 NEPOOL Load Response Program: Emissions Impacts & Associated
Discussions.
47
V-5 summarizes the results of the ISO’s analysis; it also includes NESCAUM’s estimates
of NOx and PM10 emissions using the emissions factors for diesel engines larger than 600
hp shown in Table V-1. The ISO analysis does not provide an estimate of particulate
emissions, and assumes a NOx emissions factor of 21.8 lb/MWh – approximately 32%
lower than the 32 lb/MWh emissions factor for diesel engines larger than 600 hp used for
these calculations, cited in Table V-1.41 As indicated by Table V-5, participation in ISONew England’s formal demand response programs was relatively modest in 2001,
eliciting only 24 MWh of curtailment and 479 MWh of on-site generation. All of the
generation occurred in Massachusetts over a period of 6 days.
Table V-5
Emissions Estimates for 2001 New England ISO Demand Response Programs
State
Connecticut
Massachusetts
Maine
Total
MWh
Curtailed
17.78
6.48
24.26
MWh
Generated
NOx
(tons)*
PM10
(tons)
VOC
(tons)
478.99
7.67
0.23
0.71
478.99
7.67
0.23
0.71
* New England ISO's August 2002 Report (see Footnote 40) uses a lower NOx emissions factor of 21.8
lb/MWh, resulting in total NOx emissions of 5.22 tons.
Results for Summer 2002
Between 2001 and 2002, the New England ISO substantially expanded participation in its
demand response programs, particularly with respect to the contribution from load
curtailment which grew from 24 MWh in the summer of 2001 to more than 430 MWh in
the summer of 2002. However, distributed generation also increased by more than 25%,
from 479 MWh in 2001 to more than 600 MWh in 2002. All of this curtailment and
generation was provided under the price response program, which was triggered for a
total of 166 hours on 12 different days during 2002 when market clearing prices rose
above $100/MWh. However, the ISO never found it necessary to call for voltage
reductions in 2002, so the conditions necessary to trigger the emergency demand
response program did not occur.
Table V-6 summarizes data on the number of customers who signed up to participate in
the New England ISO’s two demand response programs in 2002. Overall, 221 customers
enlisted in one or the other of these programs, providing a combined demand response
capability of 185 MW. Note that this figure for total demand response capability includes
both load curtailment capacity and on-site generation capacity (information on the
generator component of total capability was not readily discernible from the 2002 data
provided by the ISO). Because the ISO made a particular effort to enlist participants in
41
Note that the ISO analysis also accounts for offsetting emissions reductions at grid-connected peaking
units, based on the assumption that these units operated less as a result of customer-based demand
response. These or other potentially offsetting emissions benefits of on-site generation (such as reduced
demand for central-station generators to be operating in a stand-by capacity for system reliability purposes)
were not evaluated as part of this study. These issues are being addressed in the more detailed emissions
impact analysis being sponsored by EPA and conducted by Synapse Energy Economics (see Footnote 31).
48
the transmission-constrained southwest Connecticut load pocket, a disproportionate
percentage of the overall demand response capability signed up for these programs was
concentrated in that part of New England.42 As Table V-6 indicates, more customers (and
more MW) were enrolled in the emergency demand response program in southwest
Connecticut; whereas there was greater participation in the price response program
compared to the emergency program elsewhere in New England. The average demand
response capacity signed up by participants in southwest Connecticut was also
significantly greater than elsewhere in the region (an average of 2.5 MW compared to 0.5
MW).
Table V-6
New England ISO 2002 Program Enrollment43
Program Enrollment
Emergency Response
Price Response
Total
Southwest CT
Customers
MW
39
91
13
7
52
98
Other New England
Customers
MW
38
21
131
66
169
87
Total
Customers
MW
77
112
144
73
221
185
Table V-7 summarizes estimated emissions associated with the 2002 price response
program using information provided by the New England ISO and the emissions factors
for larger than 600 hp diesel engines shown in Table V-1. Table V-8 shows how on-site
generation and emissions were distributed over the 12 individual price response events in
2002, and the share of total generation and emissions attributable to participants in
different states for each event. Unlike 2001, most of the on-site generation that occurred
under the New England ISO’s formal demand response programs in 2002 occurred in
Connecticut. Finally, Table V-9 shows emissions impacts associated with on-site
generation in 2002 by price response program participants in southwestern Connecticut,
specifically.
42
Note that “southwest Connecticut” in the context of the ISO-New England programs encompasses a
somewhat larger area than Fairfield County, where PSR conducted detailed telephone surveys. However,
Fairfield County covers a large portion of southwest Connecticut, including 22 towns out of the 51 towns
that are considered to be within the transmission-constrained southwest Connecticut load pocket.
43
This table was a part of a presentation “Making Demand Response Work in New England” by Henry
Yoshimura of NE ISO in January, 2003, for the Northeast Energy and Commerce Association, available at:
http://www.ksg.harvard.edu/hepg/Papers/Yoshimura_NE.Demand.Response_01-09-2003.pdf.
49
Table V-7
Emissions Estimates for 2002 New England ISO Price Response Program
State
Connecticut
Massachusetts
Maine
New Hampshire
Rhode Island
Vermont
Total
MWh
Curtailed
272.25
88.38
62.55
7.48
3.44
0.04
434.13
MWh
Generated
421.97
27.07
81.82
0.00
0.00
72.53
603.39
NOx
(tons)*
6.76
0.43
1.31
0.00
0.00
1.16
9.67
PM10
(tons)
0.20
0.01
0.04
0.00
0.00
0.03
0.28
VOC
(tons)
0.62
0.04
0.12
0.00
0.00
0.11
0.89
* As noted in Table V-5, the New England ISO used a lower NOx emission factor in its analysis of the 2001
programs.
Table V-8
Distribution of 2002 NE ISO Price Response Events and Emissions, by State
Date
6/26/2002
7/3/2002
7/23/2002
7/30/2002
7/31/2002
8/5/2002
8/13/2002
8/14/2002
8/15/2002
8/19/2002
9/10/2002
9/16/2002
Total
MWh On-Site
Generation
3.32
4.9
60.91
37.05
16.12
31.33
11.06
130.34
221.71
13.61
47.88
25.16
603.4
NOx
(tons)
0.05
0.08
0.98
0.59
0.26
0.50
0.18
2.09
3.55
0.22
0.77
0.40
9.67
PM10
(tons)
0.00
0.00
0.03
0.02
0.01
0.01
0.01
0.06
0.10
0.01
0.02
0.01
0.28
VOC
(tons)
0.00
0.01
0.09
0.05
0.02
0.05
0.02
0.19
0.33
0.02
0.07
0.04
0.89
% of Gen % of Gen % of Gen % of Gen
in CT
in MA
in ME
in VT
27%
62%
11%
0%
93%
3%
4%
0%
16%
12%
12%
60%
17%
7%
34%
42%
48%
0%
50%
2%
67%
7%
23%
3%
54%
12%
26%
8%
82%
1%
15%
2%
91%
1%
6%
2%
50%
11%
16%
23%
83%
9%
7%
0%
38%
10%
15%
38%
70%
4%
14%
12%
Table V-9
Generation and Estimated Emissions for 2002 NE ISO Price Response Program in
Southwest CT
Date
MWh Generated
06/26/02
07/03/02
07/23/02
07/30/02
07/31/02
08/05/02
08/13/02
08/14/02
08/15/02
08/19/02
09/10/02
09/16/02
Total
0.29
2.52
6.99
4.57
0.37
2.91
2.02
2.16
2.14
1.47
0.80
2.14
28.36
NOx
(tons)
0.0046
0.0404
0.1124
0.0734
0.0060
0.0468
0.0325
0.0347
0.0344
0.0236
0.0128
0.0345
0.456
50
PM10
(tons)
0.0001
0.0012
0.0033
0.0021
0.0002
0.0014
0.0009
0.0010
0.0010
0.0007
0.0004
0.0010
0.013
VOC
(tons)
0.0004
0.0037
0.0103
0.0067
0.0006
0.0043
0.0030
0.0032
0.0032
0.0022
0.0012
0.0032
0.042
The price response program in New England is relatively new and hence participation to
date has been modest. However, New England ISO has indicated that it hopes to
significantly expand participation in this program in the future, while also continuing a
robust emergency demand response program. Ultimately, the ISO’s objective is to
increase the total demand response capability available to the pool in emergency or high
price situations to as much as 600 MW. Achieving this objective amounts to tripling the
current capacity enlisted in these programs. If that were achieved and if the price
response and/or emergency response programs were triggered more frequently in future
summers, the associated emissions impacts could increase accordingly. This could also
be compounded by an increase in participation from customers who are already signed up
for the price response program. During 2001 and 2002, many customers were signed up
but chose to not officially operate their engines under the program.
In general, this preliminary analysis suggests that additional emissions impacts associated
with the use of stationary diesel generators in recently introduced formal demand
response programs have to date been small – on an annual basis – relative to state and
local inventories of emissions from all pollutant sources. Moreover, emissions from the
existing operation of larger, non-emergency engines for peak-shaving and baseload
purposes (as described in Section B of this chapter) are likely to dwarf any near-term
increase in emissions associated with the use of diesel generators under the formal
demand response programs being introduced or augmented by grid operators. In New
Hampshire, for example, no new generation occurred under the New England ISO’s
formal price response program in 2002. However, the New Hampshire Department of
Environmental Services had documented a significant increase in NOx emissions from
stationary IC engines in the 1990s, presumably as the result of the increased operation of
non-emergency engines in response to other market factors.44
D. Potential Impact of Real-Time Pricing on Future Emissions from
On–Site Generators
Another source of impetus for increased reliance on distributed generators in coming
years – and one which may prove much more difficult to track than participation in
formal demand response programs45 – is the potential exposure of growing numbers of
44
Specifically, NOx emissions from fossil-fuel fired IC engines grew from 1.4% to 14% of New
Hampshire’s total ozone season NOx inventory from electric generating units between 1993 and 1999. In
just the 3 years between 1996 and 1999, estimated NOx emissions from these engines more than doubled –
from 243 tons in the 1996 ozone season to 576 tons in the 1999 ozone season. Moreover, the NH
Department of Environmental Services was aware of at least two instances where multiple IC engines were
installed to reduce electricity costs. These developments prompted NH to adopt new emissions rules for IC
engines in 2001.
45
Because participants in formal demand response programs are typically offered incentives or capacity
payments, or in some cases may participate like conventional generators in day-ahead or real-time
electricity markets, the ISO will generally have, at a minimum, a record of the total demand response
provided and who provided it. By contrast, the ISO would not necessarily have information on customers
who simply choose on their own initiative to avoid high prices or capacity charges by reducing
consumption of grid-supplied electricity during peak demand periods.
51
customers to real-time electricity prices. The desirability of moving more customers to
real-time pricing in the interests of promoting a more robust and efficient electricity
market is widely recognized among analysts and economic regulators (including
FERC).46 That option is most likely to be feasible for (and attractive to) large commercial
or industrial electricity users that are in a position to respond to temporal price
fluctuations by either curtailing load or by self-generating. How much on-site generation
might occur in direct response to price signals depends on the relative economics of
operating an on-site generator versus purchasing electricity off the grid. Generally, the
costs of conventional distributed generation have been estimated to range from 7 to 15
cents per kilowatt-hour. Within that range, diesel engines tend to be among the cheapest
options for on-site generation, with combined operation and maintenance costs ranging
from approximately 6 cents/kWh for larger engines, up to 7.25 cents/kWh for smaller
engines (i.e. below 500kW).47 Though these costs are well above the average wholesale
cost of grid-supplied electricity in most parts of the country, they are well below the spot
market price spikes that have occurred with some regularity during peak demand periods
in the Northeast and elsewhere during recent years.
To get some sense of how often it might be economic to operate on-site diesel generators
instead of purchasing centrally generated electricity,48 NESCAUM examined recent spot
market prices posted by the New England and New York ISOs. The results, in terms of
the numbers of hours when each pool’s prices rose above certain levels in 2002 and in the
first three months of 2003, are presented in Table V-10.49 Figure V-1 presents the number
of hours when the price of electricity exceeded 8 cents/kWh and 10 cents/kWh in the two
regions, for each month in 2002. In New England, the hourly energy clearing price (ECP)
rose above 8 cents/kWh for 77 hours in 2002; in New York, the figure was 136 hours. In
2001, the ECP rose above 8 cents/kWh for 215 hours in New England, and for 146 hours
in New York. Interestingly, the record for January and February of 2003 shows a
surprising number of high-price hours in both ISOs for these months. This may be an
isolated phenomenon attributable to the region’s unusually cold weather and high energy
prices in other sectors during these months, but it does suggest that both power control
46
At present, the vast majority of customers – and nearly all small customers – pay a flat, fixed rate for
each kilowatt-hour regardless of when they consume that kilowatt-hour. As such they are impervious to the
fact that the real cost of supplying a marginal kilowatt-hour can fluctuate dramatically from hour to hour.
The non-existence of any real-time price signal that would allow demand to respond to supply has been
identified as a major shortcoming of many current electricity markets.
47
The 7-15 ¢/kWh cost range for all forms of conventional distributed generation is taken from an Arthur
D. Little White Paper: Distributed Generation: Understanding the Economics, 1999. The cost figures cited
for large versus small diesel generators come from an NRDC report, Distributed Resources and Their
Emissions: Modeling the Impacts, Greene, Hammerschlag, and Keith, 2001.
48
Of course, operating costs will vary with individual engines, as well as with fuel costs. In addition, a
number of other factors can shift the economic calculation or otherwise influence the decision to operate an
on-site generator. For example, utilities sometimes charge high “stand-by” rates precisely to discourage onsite generation, or permit restrictions may apply, or there may be an additional hassle factor (to the facility
operator) involved in switching to on-site generation.
49
Historically, the New England ISO has recorded a pool-wide Energy Clearing Price or ECP (although it
has recently also introduced location-based marginal pricing). In Table V-10, the Location Based Marginal
Price (LBMP) represents the real time prices of electricity in the New York City area.
52
areas may be subject to wintertime, as well as summertime price spikes under certain
conditions.
Table V-10
New England Energy Clearing Price (ECP) New York City Location Based
Marginal Price (LBMP) Data
Electricity Prices
2001
2002
Hours above Hours above Hours above Hours above Average
10¢/kWh
8¢/kWh
7¢/kWh
6¢/kWh
Prices
New England ECP
106
215
381
787
3.99¢
New York City LBMP
313
549
739
1066
4.47¢
New England ECP
New York City LBMP
53
256
77
557
133
890
299
1643
3.54¢
4.47¢
87
87
141
324
6.02¢
January New England ECP
2003
New York City LBMP
February New England ECP
a
154
269
392
505
7.95¢
78
179
278
401
6.94¢
8.60¢
2003
New York City LBMP
161
316
444
508
March
New England ECP
N/A
a
N/A
a
N/A
a
N/A
N/A
2003
New York City LBMP
166
286
369
470
8.16¢
a
a
value not available at the time of publication
Assuming, for example, that an additional 600 MW of diesel engine capacity in New
England or New York were to operate an average of 100 hours per year simply in
response to high short-term prices, the additional generation (60,000 MWh) and
emissions would substantially exceed – by as much as two orders of magnitude – the
amounts estimated in connection with formal demand response programs sponsored by
the NY and NE ISOs to date. Whether the above calculation grossly overstates or
understates the potential impact of real-time pricing on engine operation in New England
or New York is, of course, a different question, and one that was outside the scope of this
study to evaluate. To put the same (essentially arbitrary) figure of 60,000 MWh in a
different perspective, however, it may also be useful to point out that it is much lower
than the output estimated in earlier sections of this chapter for generators either surveyed
or registered in New York City and Fairfield County. The results of PSR’s telephone
surveys in these areas suggest that current engine output – when one includes the nonemergency peak shaving and baseload generators that account for most hours of
operation – is already on the order of hundreds of thousands of MWh per year. Whether
the (typically larger) non-emergency units are in a position to substantially increase their
output in response to stronger price signals would therefore be an important question to
investigate.50
50
At present, emergency engines in the Northeast are almost without exception precluded from operating
in response to real-time price signals by current permitting restrictions. Whether these permitting
restrictions are always respected, and how effectively they can be enforced, is a separate but important
question for state regulators and policymakers.
53
Figure V-1
Hours that Electricity Prices Exceeded 8¢ and 10¢/kWh in 2001 and 2002
New England Energy Clearing Price 2002
New England Energy Clearing Price 2001
Add'l hours
above 8¢/kWh
Hours above
10¢/kWh
80
60
40
20
40
Hours above price
30
Add'l hours
above 8¢/kWh
Hours above
10¢/kWh
20
10
Dec
Nov
Oct
Sep
Aug
Jul
60
40
20
Month of price data
54
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Month of price data
Apr
0
Jan
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
0
80
Mar
20
120
100
Feb
40
Feb
Add'l hours
above 8¢/kWh
Hours above
10¢/kWh
140
Hours above price
60
Jan
Hours above price
NYC Location Based Marginal Price 2002
160
Add'l hours
above 8¢/kWh
Hours above
10¢/kWh
80
Jun
Month of price data
NYC Location Based Marginal Price 2001
100
May
Month of price data
Apr
Jan
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
Mar
0
0
Feb
Hours above price
100
VI. Emission Control Technologies for Stationary Diesel
Generators
The operation of all internal combustion (IC) engines results in the emission of
hydrocarbons (HC or VOC51), carbon monoxide (CO), nitrogen oxides (NOx), and
particulate matter (PM). The actual concentrations of these pollutants in engine exhaust
vary from engine to engine, depend on the mode of operation and are strongly related to
the type of fuel used. An important emissions control priority for diesel engines
(compared to stationary IC engines using other fuels such as natural gas, propane or
gasoline) is particulate matter. Indeed, for the reasons discussed in the previous chapter,
particulate emissions pose perhaps the most serious public health risks associated with
diesel exhaust.
Various emission control technologies exist for IC engines, which can afford substantial
reductions in all four pollutants listed above. Uncontrolled emissions and achievable
emissions reductions depend on fuel characteristics, on whether the engine is being run
rich, lean, or stoichiometrically52 and on the emission control technology being used. The
most important control options include oxidation catalysts, which can be used to control
HC, CO and PM emissions; selective catalytic reduction or SCR, which is very effective
for NOx control; and particulate filters, which can achieve high levels of particulate
control. Other emissions control options discussed in this chapter include lean-NOx
catalysts, ignition timing adjustments and exhaust gas recirculation (EGR). Importantly,
both filter and catalyst controls can provide significant (80-90%) reductions in HC
emissions – including emissions of toxic HC – in addition to PM and CO reductions.
Filters can also help eliminate the characteristic odor of diesel engine exhaust. Finally, an
important issue in assessing control options for diesel IC engines is fuel quality, and
sulfur content in particular. All of the major control options discussed in this report
perform significantly better with – and in some cases require – lower sulfur fuel.
The discussion of stationary IC engine control technologies in this chapter was prepared
by ESI International, Inc. for NESCAUM as a complement to the engine inventory and
related efforts described elsewhere in this report. It begins by providing some additional
background on the status and market for distributed generation technology generally.
(Additional background information on stationary generator engines is provided in
Appendix D.) This introductory section is followed by a summary of emissions factors
for uncontrolled engines and a review of available control options for diesel and natural
gas-powered stationary engines used as distributed power generation resources. Chapter
51
The acronyms correspond to “hydrocarbons” (HC) and “volatile organic compounds” (VOC). See
additional explanation for these terms in Footnote 27.
52
In stoichiometric combustion, the air-fuel mixture is theoretically balanced such that there is exactly
enough oxygen to allow for complete oxidation of the fuel. In an engine running rich, the air-fuel ratio is
tipped toward relatively more fuel and less air, whereas in an engine running lean the ratio is tipped toward
relatively less fuel and more air. All diesel engines inherently operate lean, whereas natural gas fired
engines, for example, can operate in all three modes.
55
VII, also prepared by ESI International, Inc., provides six detailed case studies of actual
control technology installations on diesel generators.
A. Distributed Generation Market Status and Outlook
As noted elsewhere in this report, stationary reciprocating IC engines are the most
common and most technically mature of all distributed generation technologies. They are
available from small sizes (e.g. 5 kW for residential back-up generation) to large
generators (e.g. 7 MW). Stationary diesel engines are, by a large margin, the most
commonly used engines for distributed generation. In fact, over 90% of distributed
generators are powered by diesel fuel.53 In 2001, the California Air Resources Board
(CARB) surveyed stationary IC engines throughout California and found that dieselpowered Caterpillar engines were by far the most commonly used engine, as shown in
Figure VI-1.
Figure VI-1
Permitted Stationary Diesel Engine Database Manufacturer Listing
Source: California Air Resources Board “Permitted Stationary Diesel Engine Database.”
January 16, 2002.
In recent years, the convergence of a number of factors has led to a growing national
interest in distributed generation generally. They are:
•
•
A need for the existing electric power infrastructure to keep pace with demand
for high-quality, reliable power;
Dramatic reduction in large electric generating plant investment due to
regulatory, capital, environmental and political constraints;
53
U.S. Department of Energy, Office of Distributed Energy Resources. Web Site: Distributed Energy
Resources, www.eren.doe.gov/der.
56
•
•
Restructuring of the power industry leading to competitive markets and
reduced incentives for utilities to invest in new generating facilities; and
Technological advancements in small-scale power generating equipment with
greater efficiencies, improved environmental performance and lower costs.
1. National Market
Nationally, many large industrial customers have already embraced distributed
generation. The largest potential new markets for distributed generation in the near term
therefore lie with commercial and small to medium-sized industrial customers. Currently,
there are more than 60,000 MW of distributed power installed in North America in the
form of IC engines and gas turbines.54 Their use in providing back-up power continues to
grow steadily at 7% per year. Other distributed generation applications for meeting
baseload and peaking requirements are growing at 11% and 17% respectively.55
A number of trends in distributed generation markets are now becoming evident. There is
increasing use of:
•
•
•
•
•
•
On-site power for cost and reliability reasons.
Combined heat and power (CHP) or cogeneration applications.
Combinations of CHP with thermal energy storage systems.
Fuel cells.
Uninterruptible power systems to ride through brief power disturbances and
outages.
Increased use of natural gas for distributed generation. (In fact, DOE has
estimated that distributed generation using natural gas could account for as
much as 20% of all power generation nationwide by the year 2020.)
2. Small Commercial and Residential Markets
Advances in new generating technologies have been moving in the direction of smaller
equipment with increased output, making on-site generation increasingly feasible for
commercial establishments and even residential energy users. Small manufacturing plants
and medium-sized buildings can now be powered by cost-effective combustion turbines
as small as 500 kW, while IC engines have become cost-effective for systems down to 50
kW, making them suitable for small office buildings and even for free-standing
commercial establishments such as restaurants.
54
As indicated in the Introduction to this report, other estimates of installed distributed generation capacity
nationwide are even higher, ranging up to 127,000 MW using PSR’s estimation methodology.
55
U.S. Department of Energy, Office of Distributed Energy Resources. Web Site: Distributed Energy
Resources, www.eren.doe.gov/der.
57
In the residential market, utilization of distributed generation technologies is growing
among affluent U.S. consumers. A recent national survey found measurable interest in
customer-site generation within this group:56
•
•
•
•
•
•
More than half reported their power usage had increased over the past five
years.
31% expressed interest in generating on-site power.
Nearly 10% had purchased or leased an emergency or back-up generator for
their primary residence.
16% used such equipment in weekend or vacation homes.
40% were considering purchasing a generator.
Congress is also becoming interested in this issue. The Home Energy
Generation Act, introduced in 1999, is aimed at setting standards to encourage
the development of residential distributed energy technologies.
3. Combined Heat & Power Market
A thriving U.S. market exists for IC engine-based combined heat and power (CHP)
applications (see Footnote 11). The top twelve states with CHP sites using stationary IC
engines are shown in Table VI-1 below.
Table VI-1
Top Twelve States with Combined Heat & Power Sites Using Stationary IC Engines
State
California
New York
New Jersey
Massachusetts
Illinois
Pennsylvania
Connecticut
Michigan
Texas
Virginia
Florida
Arizona
Total
Number of Sites
493
136
117
46
44
44
43
28
23
17
16
15
1022
Percent of U.S. Market
42.0
11.6
10.0
3.9
3.7
3.7
3.7
2.4
2.0
1.4
1.4
1.3
87.1
Source: Resource Dynamics Corporation. Web Site: Distributed Generation Information Center,
www.distributed-generation.com.
56
Resource Dynamics Corporation. Web Site: Distributed Generation Information Center,
www.distributed-generation.com.
58
4. Utility Market
Across the nation, the use of distributed generating resources among utility companies is
gaining momentum. Some recent examples include:
•
•
•
•
•
•
Tacoma Public Utilities in Washington State installed 30 diesel-fired engines
for a combined capacity of 48 MW.
GPU Energy proposed a program to New Jersey regulators for the dispatch of
customer-owned diesel stand-by generators, with the aim of overcoming
delivery system bottlenecks in certain counties.
PacifiCorp leased and installed five 22 MW gas turbines adjacent to the main
office of its subsidiary, Utah Power, in Salt Lake City.
The New York Power Authority is seeking regulatory approval to install
eleven gas turbines (with a collective capacity of 444 MW) at six sites in New
York City boroughs and on Long Island.
Long Island Power Authority has added a new supplemental service rate to
on-site generators that will provide lower demand charges during the summer
and lower energy charges throughout the year.
Transmission grid operators (Independent System Operators or ISOs) are
establishing programs to reward energy users who shed utility load by using
on-site generators. Programs established by PJM,57 ISO-New England and the
New York ISO are the most favorable to distributed generation.
Note that utility sites for distributed generation often use several large engines in parallel
to produce larger amounts of electricity, as illustrated in Figure VI-2.
Figure VI-2
An Example of Multiple Engine Usage (Tacoma Power, CAT 3516B Diesel Engines
with SCR)
SCR Catalyst
Engine Housing
57
See description of PJM in Footnote 8.
59
B. Emissions Factors for Stationary IC Engines
As noted previously, emissions vary from engine to engine and model to model and
depend on mode of operation as well as fuel qualities. Nonetheless, the values shown in
Table VI-2 are representative of what may be expected for typical engines. Note that the
NOx and PM10 emissions factors shown in Table VI-2 are consistent with those used to
estimate emissions impacts in Chapter V of this report (see Table V-1).
Table VI-2
Typical Emissions Factors for Stationary IC Engines
Fuel Type Engine Type
Diesel
Gasoline
Natural
Gas
a
Up to 600 hp
Greater than 600 hp
All
2-Cycle Lean Burn
4-Cycle Lean Burn
4-Cycle Rich Burn
NOx
14.06
10.86
4.99
10.89b
11.79b
9.98b
Emission Factor (g/hp-hr)
CO
VOC
SOx
8.5
1.14
0.1645
8.5
1.0b
0.1645
199.13
9.79
0.1645
1.5
0.43
0.002
b
1.6
0.721
0.002
8.62
0.14b
0.002
PM10a
1.00
0.318
0.327
0.152b
0.152b
0.152b
Although measured as PM10, particles smaller than 2.5 microns (PM2.5) account for most of the overall
PM emissions from IC engines.
b
These emission factors were published by the SMAQMD based on results obtained on existing engines
in the course of testing for BACT determinations. They reflect higher emissions rates than the AP-42
factors (see Footnote 36).
Emissions factors such as those summarized in Table VI-2 are important for developing
emission control strategies, determining applicability of permitting and control programs,
understanding the effects of sources and appropriate mitigation strategies, and a number
of other related applications by various users, including federal, state and local agencies,
and industry. For obvious reasons, data from source-specific emission tests or continuous
emission monitors are generally preferred for estimating actual emissions from individual
sources. Unfortunately, source-specific test data are not always available; hence,
emission factors are frequently used to estimate emissions, in spite of their limitations.
C. Emission Control Technologies for Diesel and Natural Gas
Stationary IC Engines Used in Distributed Power Generation
Most engines used for distributed power generation are powered by either diesel or
natural gas. The main emissions control priority for diesel-powered engines has been PM
and NOx emissions. For natural gas-powered engines, the control priority has been NOx
emissions.
As noted in the introduction to this chapter, PM emissions from diesel-powered engines
can be controlled with both diesel particulate filters (DPFs) – which provide very high
(>80%) reductions – and diesel oxidation catalysts (DOCs) for more modest (>25%)
60
reductions. Both DPFs and DOCs can also provide significant reductions in HC, CO and
toxic HC emissions. They also serve to eliminate the characteristic odor of diesel engines.
NOx emissions, from both diesel and lean-burn natural gas-powered stationary IC
engines, have traditionally been controlled using SCR. Here, a reagent – typically
ammonia or urea – is added to the oxygen-rich exhaust environment and reacted over a
catalyst with the NOx present in the exhaust gases. SCR systems commonly achieve NOx
control efficiencies in excess of 90%. Lean-NOx catalysts can also be used for more
modest control (15-25%) of NOx emissions from IC engines. Incorporating an oxidation
catalytic function with these technologies also allows for the simultaneous control of CO,
HC and toxic HC emissions.
Emissions from rich-burn natural gas powered engines can be controlled using nonselective catalytic reduction (NSCR) which can achieve NOx reductions in excess of 90%
as well as significant HC reductions. Three-way catalyst (TWC) technology – similar to
automotive catalyst technology – can be used on stoichiometrically calibrated natural gas
engines for the simultaneous control of NOx, CO, HC and toxic HC emissions.
Figure VI-3 illustrates typical gaseous emissions control performance at various air/fuel
ratios for the above mentioned catalysts with respect to CO, NOx and non-methane
hydrocarbon (NMHC)58 emissions.
Figure VI-3
Control Capabilities of Catalyst Control Technology vs. A/F Ratio
SCR /Lean-NOx
58
See Footnote 27 for further discussion of the term NMHC.
61
Figure VI-3 (cont.)
Control Capabilities of Catalyst Control Technology vs. A/F Ratio
SCR /Lean-NOx
SCR
NSCR
Lean-NOx
Source: Manufacturers of Emission Controls Association, “Emission
Control Technology for Stationary Internal Combustion Engines –
Status Report,” July 1997.
62
Where diesel particulate filters are used for PM control, an oxidizing catalyst function is
often included. When this function is incorporated into a diesel particulate filter, CO and
hydrocarbon emissions can be controlled as shown above for oxidation catalysts.
The vast majority of stationary IC engines used for distributed power generation are leanburn diesel engines. Hence, the rest of this chapter focuses on the following technologies:
•
•
•
Diesel particulate filters for control of PM and other gaseous emissions from
diesel engines,
Oxidation catalysts for control of PM and other gaseous emissions from diesel
engines, and
Selective catalytic reduction (SCR) for the control of NOx and other gaseous
emissions from diesel engines.
To date, the use of emission control technology on stationary diesel engines has been
limited. Nevertheless, hundreds of control technology installations exist, with most
involving retrofits to existing engines. In addition, control technologies have been
demonstrated at installations in Europe, Korea, and Taiwan.
Other emissions control options applicable to diesel engines – such as crankcase emission
control, EGR and lean-NOx catalysts – are described later in this chapter.
1. Diesel Particulate Filters (DPFs)
As the name implies, diesel particulate filters remove particulate matter in diesel exhaust
by filtering exhaust from the engine. They can be installed on both stationary and mobile
engines (e.g. diesel vehicles). Since a filter can fill up over time, these systems must
provide a means of burning off or removing accumulated particulate matter. A
convenient means of accomplishing this is to burn or oxidize accumulated particulate
matter on the filter when exhaust temperatures are adequate. By burning off trapped
material, the filter is cleaned or “regenerated.” Filters that regenerate in this fashion
cannot be used in all situations. Both exhaust gas temperature and fuel sulfur level must
be taken into consideration.
In some non-road applications, disposable filter systems have been used. A disposable
filter is sized to collect particulates for a working shift or some other predetermined
period of time. After a prescribed amount of time, or when backpressure limits are
approached, the filter is removed and cleaned or discarded. To ensure proper operation,
disposable filter systems must be designed for specific engines and engine applications.
In many cases, the use of these types of filters on stationary diesel engines for distributed
power generation may not be practical.
A number of filter materials have been used in diesel particulate filters, including:
ceramic and silicon carbide materials, fiber wound cartridges, knitted silica fiber coils,
ceramic foam, wire mesh, sintered metal substrates and temperature resistant paper in the
63
case of disposable filters. Collection efficiencies of these filters range from 50% to over
90 %. Filter materials capture particulate matter by interception, impaction and diffusion.
Filter efficiency has rarely been a problem with the filter materials listed above, but work
has continued to: (1) optimize filter efficiency and minimize back pressure, (2) improve
the radial flow of oxidation in the filter during regeneration, and (3) improve the
mechanical strength of filter designs. Figure VI-4 provides a diagram of a typical
wallflow-type filter system.
Figure VI-4
Mechanical Filtration of a Wallflow DPF
Intake Æ
Æ Exhaust
As shown in Figure VI-4, particulate-laden exhaust enters the filter from the left. Because
the cells of the filter are capped at the downstream end, exhaust cannot exit the cell
directly. Instead, exhaust gas passes through the porous walls of the filter cells. In the
process, particulate matter is deposited on the upstream side of the cell wall. Cleaned
exhaust gas exits the filter to the right.
Many techniques can be used to regenerate a diesel particulate filter. Some of these
techniques are used together in the same filter system to achieve efficient regeneration.
Both on- and off-board regeneration systems exist. The major regeneration techniques are
listed below.
•
Catalyst-based regeneration using a catalyst applied to the surfaces of the
filter. A base or precious metal coating applied to the surface of the filter
reduces the ignition temperature necessary to oxidize accumulated particulate
matter.
•
Catalyst-based regeneration using an upstream oxidation catalyst. In this
technique, an oxidation catalyst is placed upstream of the filter to facilitate
oxidation of nitric oxide (NO) to nitrogen dioxide (NO2). The nitrogen dioxide
adsorbs on the collected particulate, substantially reducing the temperature
required to regenerate the filter.
•
Fuel-borne catalysts. Fuel-borne catalysts reduce the temperature required for
ignition of trapped particulate matter.
64
•
Air-intake throttling. Throttling the air intake to one or more of the engine
cylinders can increase the exhaust temperature and facilitate filter
regeneration.
•
Post top-dead-center fuel injection. Injecting small amounts of fuel in the
cylinders of a diesel engine after pistons have reached the top-dead-center
position introduces a small amount of unburned fuel in the engine’s exhaust
gases. This unburned fuel can then be oxidized in the particulate filter to
combust accumulated particulate matter.
•
On-board fuel burners or electrical heaters. Fuel burners or electrical heaters
upstream of the filter can provide sufficient exhaust temperatures to ignite
accumulated particulate matter and regenerate the filter.
•
Off-board electrical heaters. Off-board regeneration stations combust trapped
particulate matter by blowing hot air through the filter system.
Experience with catalyst-based filters indicates that they can achieve a virtually complete
reduction in odor and in the soluble organic fraction of particulate emissions. However,
some catalysts may also increase sulfate emissions.
Sulfur in diesel fuel affects the reliability, durability and emissions performance of
catalyst-based diesel particulate filters. The degree of the impact is dependent both on the
technology being applied and the operation of the engine. Sulfur affects filter
performance by inhibiting the performance of catalytic materials upstream of, or on the
filter. Sulfur also competes with chemical reactions intended to reduce pollutant
emissions and creates particulate matter through catalytic sulfate formation. In general,
the less sulfur in the fuel, the better the control technology performs. Therefore, when
using diesel particulate filter technology, a careful assessment of its suitability should be
made based on fuel sulfur content, engine type, filter system, operating conditions and
desired control levels.
Filter systems do not appear to cause any additional engine wear or affect engine
maintenance. The systems are designed with engine displacement as a key parameter in
order to ensure that appropriate backpressures are encountered during engine operation.
Concerning maintenance of the filter system itself, manufacturers are designing systems
to minimize maintenance requirements during the useful life of the engine. In some cases,
however, accumulated lubricating oil ash may have to be periodically removed.
Generally, manufacturers provide the end-user with appropriate removal procedures.
Determining whether a given filter system is appropriate for a given engine in a specific
distributed power generation application depends on fuel sulfur level and exhaust gas
temperature during operation. The steady-state operation of these engines makes this
determination relatively simple. For example, a manufacturer may specify that, when
using diesel fuel containing 50 ppm of sulfur, regeneration of the filter will occur at
temperatures in excess of 300oC. Running the engine at the load condition in which it will
65
be used to generate electricity and measuring the exhaust gas temperature will indicate
whether the filter system is suitable.
Generally, the emissions control performance of diesel particulate filters is well
established. While most emissions testing has been performed on transient test cycles,
steady state test data also exist.59 Figure VI-5 shows control performance for a diesel
particulate filter applied to a 1998 model year, 400 hp engine under the 13 different
operating modes described in Table VI-3. The sulfur content of the fuel used in these
tests was 54 ppm (parts per million). As can be seen in Figure VI-5, significant PM
reductions were achieved in all operating modes.
Table VI-3
Specifications of a 13-Mode Engine Test Cycle
Mode
1
2
3
4
5
6
7
8
9
10
11
12
13
Speed, rpm
600
840
840
1080
1080
1380
1380
1560
1560
1560
1800
1800
1800
Torque, %
0
50
100
50
100
50
100
50
75
100
50
75
100
Figure VI-5
13-Mode Steady-State Test Cycle for DPF PM Emissions Reduction
(g/bhp-hr)
0.03
0.025
0.02
Baseline
0.015
DPF
0.01
0.005
0
1
59
2
3
4
5
6
7 8
Mode
9
10 11 12 13
Manufacturers of Emission Controls Association, Stationary Engine Emission Control, May 2002.
66
The use of ultra-low (e.g. <15 ppm) sulfur fuel would have resulted in even more
dramatic reductions by virtually eliminating the catalytic production of sulfuric acid.
Also, formulating the catalyst to minimize the production of sulfuric acid would have
further enhanced its control capabilities.
Another advantage of applying particulate filters to diesel-powered stationary IC engines
is their ability to dramatically reduce toxic hydrocarbon emissions. As part of the same
test program used to generate the results shown in Figure VI-5, the control capability of
two separate, catalyst-based DPF systems was evaluated with respect to 18 distinct
polycyclic aromatic hydrocarbons (PAHs). The results are shown in Table VI-4. As
indicated, average PAH emissions were reduced by 89% and 84% for the systems tested.
In this case, the testing was performed over the U.S. Federal Test Procedure (FTP), a
transient test cycle used for motor vehicles. Test results would likely be similar, or better,
on a steady-state test cycle more representative of a stationary IC engine because of the
relatively low load associated with parts of the FTP and correspondingly low exhaust gas
temperatures (the catalyst function of the filter performs better at elevated temperatures).
Table VI-4
Reductions in PAH Emissions for Two DPF Systems (micrograms per bhp-hr)
Compound
Baseline DPF-A DPF-B % Red DPF-A % Red DPF-B
295
50
0
83.0%
100.0%
Napthalene
635
108
68
83.0%
89.3%
2-Methylnapthalene
40
0.8
1
98.0%
97.5%
Acenapthalene
46
6.7
11
85.4%
76.1%
Acenapthene
72
29
12
59.7%
83.3%
Fluorene
169
33
26
80.5%
84.6%
Phenanthrene
10
1
1
90.0%
90.0%
Anthracene
7.7
0
2
100.0%
74.0%
Fluoranthene
14
0
2
100.0%
85.7%
Pyrene
0.22
0
0.01
100.0%
95.4%
Benzo(a)anthracene
0.51
0
0
100.0%
100.0%
Chrysene
0.26
0
0
100.0%
100.0%
Benzo(b)fluoranthene
0.15
0
0
100.0%
100.0%
Benzo(k)fluoranthene
0.26
0
0
100.0%
100.0%
Benzo(e)pyrene
0.01
0
0
100.0%
100.0%
Perylene
0.13
0
0
100.0%
100.0%
Indeno(123-cd)pyrene
0.01
0
0
100.0%
100.0%
Dibenz(ah)anthracene
0.32
0
0
100.0%
100.0%
Benzo(ghi)perylene
Average Reduction
89%
67
84%
2. Diesel Oxidation Catalysts (DOCs)
The principle behind oxidation catalysts is that they reduce emissions by causing
chemical reactions without themselves being changed or consumed. A DOC system
typically consists of a steel housing that contains a metal or ceramic structure which acts
as a catalyst support or substrate. The size of the catalyst system depends on the size of
engine to which it is being applied. There are no moving parts, just acres of interior
surfaces on the substrate coated with either base or precious catalytic metals such as
platinum (Pt), rhodium (Rh) and palladium (Pd). Catalysts transform pollutants into
harmless gases by causing chemical reactions in the exhaust stream. Diesel oxidation
catalysts serve to reduce PM, CO, HC and toxic HC emissions. PM emissions are
reduced by the chemical transformation of their soluble organic fraction to carbon
dioxide and water. Figure VI-6 outlines the functionality of a diesel oxidation catalyst.
Figure VI-6
Diesel Oxidation Catalysts
CO
Aldehyde
HC
PAH
CO2
H2O
SO2
/SO3
SO2
NOx
NO
C2H2n
PA
So
Meta
SO2+H
Flow through
monolith
CO + 1/2
O2
HC + O2
PAH + O2
SO2+H
CO2
CO2 +
H2O
CO2 +
So
Meta
Different catalyst formulations can be used to target different pollutants more
aggressively than others, as illustrated by a comparison of different catalyst formulations
(“C” & “E”, respectively) as part of the previously-referenced test program. The results
of this comparison over the FTP test cycle are graphed in Figure VI-7.
As indicated by Figure VI-7, catalyst formulation “C” was designed to aggressively
reduce both HC and CO, whereas it achieved only modest PM reductions. On the other
hand, formulation “E” reduced PM emissions significantly more at the expense of some
HC and CO control. These results illustrate why it is important to select an oxidation
catalyst knowing which pollutants are of most concern.
68
Figure VI-7
DOC Performance
(g/bhp-hr)
0.14
0.12
0.1
Baseline
DOC C
DOC E
0.08
0.06
0.04
0.02
0
HC
CO/10
PM
Pollutant
As part of the same test program, different oxidation catalysts were also tested over the
steady-state test cycle outlined in Table VI-3. The results of these tests are shown in
Figure VI-8. It indicates that the oxidation catalyst was effective in reducing PM
emissions in all modes except mode 5 (no change) and mode 3 where there was a slight
increase in PM emissions. Significant reductions were found in all other modes – in some
instances exceeding 50%. This testing was performed using fuel with a sulfur content of
368 ppm. As is the case with catalyst-based particulate filters, the use of ultra-low (e.g.
<15 ppm) sulfur fuel would have resulted in greater reductions and certainly would have
eliminated the PM emissions increase found in mode 3 operation.
Like catalyst-based DPFs, oxidation catalysts are effective in controlling toxic HC
emissions. Table VI-5 outlines the control capabilities of two DOCs on the same 18
distinct PAHs included in Table VI-4. As indicated, reductions in excess of 50% are
readily achieved using the FTP, with reductions approaching 70% for some compounds.
For the reasons indicated above, similar or better results can be expected on steady-state
test cycles.
Diesel oxidation catalysts are virtually maintenance-free. Periodic inspection to ensure
that cell plugging is not occurring is advisable. This would only occur in the case of
engine malfunction (e.g. a faulty injector or two) and in that event, the catalysts can
easily be cleaned and reinstalled.
69
Figure VI-8
13-Mode Steady-State Test Cycle for DOC PM Emissions Reductions
(g/bhp-hr)
0.035
0.03
0.025
0.02
Baseline
DOC
0.015
0.01
0.005
0
1
2
3
4
5
6
7
8
9
10 11 12 13
Mode
Like diesel particulate filters, diesel oxidation catalysts are also affected by sulfur. Hence,
the sulfur content of diesel fuel is critical to applying catalyst technology. Catalysts used
to oxidize the soluble organic fraction of particulate emissions can also oxidize sulfur
dioxide to form sulfates, which are counted as part of total particulate emissions. This
reaction is not only dependent on the level of sulfur in the fuel, but also on the
temperature of the exhaust gases. Catalyst formulations have been developed which
selectively oxidize the soluble organic fraction while minimizing oxidation of the sulfur
dioxide. However, the lower the sulfur content in the fuel, the greater the opportunity to
maximize the effectiveness of oxidation catalyst technology for both better total control
of PM and greater control of toxic hydrocarbons. The lower, 500 ppm sulfur fuel that was
introduced in 1993 throughout the U.S. has thus facilitated the application of catalyst
technology to diesel-powered vehicles. Furthermore, the very low fuel sulfur content (30
ppm) available in several European countries, and more recently in the U.S., has further
enhanced catalyst performance.
Catalysts have also been installed on engines using higher sulfur fuel (e.g. >500 ppm
sulfur). The performance of an oxidation catalyst on such fuel will vary with catalyst
formulation, engine type and duty cycle. In all cases, however, catalyst performance is
adversely affected by the presence of sulfur in the fuel.
70
Table VI-5
Reductions in PAH Emissions for DOCs (micrograms per bhp-hr)
Compound
Baseline Cat B
Cat D
% Red Cat B % Red Cat D
295
159
182
46.1%
38.3%
Napthalene
635
278
277
56.2%
56.4%
2-Methylnapthalene
40
13
13.6
67.5%
66.0%
Acenapthalene
46
25
24.4
45.7%
47.0%
Acenapthene
72
29
28.9
59.7%
59.9%
Fluorene
169
54
56
68.0%
66.9%
Phenanthrene
10
2.6
2.8
74.0%
72.0%
Anthracene
7.7
2.6
4.9
66.2%
36.4%
Fluoranthene
14
5
6.4
64.3%
54.3%
Pyrene
0.22
0.05
0.18
77.3%
18.2%
Benzo(a)anthracene
0.51
0.16
0.33
68.6%
35.3%
Chrysene
0.26
0.09
0.12
65.4%
53.8%
Benzo(b)fluoranthene
0.15
0.05
0.08
66.7%
46.7%
Benzo(k)fluoranthene
0.26
0.08
0.14
69.2%
46.2%
Benzo(e)pyrene
0.01
0
0
100.0%
100.0%
Perylene
0.13
0.04
0.07
69.2%
46.2%
Indeno(123-cd)pyrene
0.01
0
0
100.0%
100.0%
Dibenz(ah)anthracene
0.32
0.1
0.22
68.8%
31.3%
Benzo(ghi)perylene
Total
1290.57 568.77 597.14
55.9%
53.7%
Average Reduction
68.5%
54.1%
3. Crankcase Emission Controls
In most existing turbocharged diesel engines, the crankcase breather is vented to the
atmosphere – often using a downward directed draft tube to prevent fouling of the
turbocharger and the resultant maintenance. While a rudimentary filter is often installed
on the crankcase breather (the vent for the oil reservoir), a substantial amount of
particulate matter is often released to the atmosphere. For diesel engines used in motor
vehicle applications, emissions through the breather may exceed 0.7 g/bhp-hr during idle
conditions, even on recent model year engines.
One solution to this problem is the use of a multi-stage filter designed to collect, coalesce
and return the emitted lube oil to the engine’s sump.60 Filtered gases are returned to the
intake system, balancing the differential pressures involved. Typical systems consist of a
filter housing, a pressure regulator, a pressure relief valve and an oil check valve. These
systems greatly reduce crankcase emissions.
60
NOx and PM Control from Heavy-Duty Diesel Engines Using a Combination of Low Pressure EGR and
Continuously Regenerating Diesel Particulate Filter, S. Chatterjee, R. Conway, S. Viswanathan,
Johnson Matthey; M. Blomquist, S. Andersson, STT Emtec. SAE Paper No. 2003-01-0048.
71
Figure VI-9
Schematic of Crankcase Emission Control
4. Selective Catalytic Reduction (SCR)
SCR has been used to control NOx emissions from stationary sources for over 15 years.
More recently, it has been applied to select mobile sources including trucks, marine
vessels, and locomotives. Applying SCR to diesel-powered engines provides substantial
reductions of NOx emissions.
Like an oxidation catalyst, the catalyst in an SCR system allows chemical reactions to
take place that would not take place during normal engine operation. Again, like an
oxidation catalyst, the SCR catalyst enables chemical reactions without being consumed
itself. Unlike an oxidation catalyst, however, a SCR system needs a chemical reagent – or
“reductant” – to convert nitrogen oxides to molecular nitrogen and oxygen in the exhaust
stream. The reductant is typically ammonia (NH3) or urea. It is added at a rate calculated
from an algorithm that estimates the amount of NOx present in the exhaust stream. The
algorithm relates NOx emissions to engine operating conditions, for example engine
revolutions per minute and load. As exhaust gases and reductant pass over the SCR
catalyst, chemical reactions occur that reduce NOx emissions from 65% to more than
90%. Where an additional oxidation function is included, HC emissions (including toxic
emissions) can be reduced from 50-90% and PM emissions can be reduced 30-50%. Like
all catalyst-based emission control technologies, SCR performance is enhanced by the
use of low sulfur fuel. Figure VI-10 is a schematic of the functionality of an SCR system.
72
Figure VI-10
Selective Catalytic Reduction
Both precious metal and base metal catalysts have been used in SCR systems. Base metal
catalysts, typically vanadium and titanium, are used for exhaust gas temperatures
between 450oF and 800oF. For higher temperatures (675oF to 1100oF), zeolite catalysts
may be used. Precious metal SCR catalysts are also useful for low temperatures (350550oF ). Note that the benefits of low sulfur fuel and the potential for sulfuric acid
formation apply also to SCR systems that use precious metals.
The same 13-mode steady-state test cycle outlined in Table VI-3 was used to test NOx
control performance for an SCR system. The results, shown in Figure VI-11, suggest
achievable reductions ranging from 65% (in mode 12) to nearly 100% (in mode 2).
Figure VI-11
13-Mode Steady-State Test Cycle for SCR NOx Emissions Reductions
(% Reduction)
100
80
60
NOx Reductions
40
20
0
1
2
3
4
5
6
7
8
9
10
11
12
13
Mode
5. Diesel Particulate Filters Combined with SCR
Recently in the U.S., stationary diesel engines used for distributed power generation have
begun to be equipped with a combination of diesel particulate filters and SCR. This
combination of control technologies was also investigated as part of the aforementioned
73
test program. In this instance, a fuel-borne catalyst was used with the particulate filters.
The testing was performed over the FTP with fuel containing 368 ppm sulfur. The results
are shown in Figure VI-12. Again, it is likely that similar results could be achieved under
steady-state conditions. As the figure indicates, HC and PM emissions were virtually
eliminated, while NOx emissions were reduced by approximately 75% and CO emissions
were reduced by almost half.
Figure VI-12
Control Performance of a Combined DPF/SCR System
(g/bhp-hr)
0.45
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
Baseline
Nondetectable
HC
SCR+DPF
CO/10
NOx/10
PM
Pollutant
6. Exhaust Gas Recirculation (EGR)
EGR on a diesel engine offers an effective means of reducing NOx emissions. Both lowpressure and high-pressure EGR systems exist, but low-pressure EGR is most suitable for
retrofit applications because it does not require engine modifications.
As the name implies, EGR involves re-circulating a portion of the engine's exhaust back
to the turbocharger inlet or, in the case of a naturally aspirated engines, to the intake
manifold. In most systems, an intercooler lowers the temperature of the re-circulated
gases. The cooled re-circulated gases, which have a higher heat capacity than air and
contain less oxygen than air, lower combustion temperature in the engine and reduce
NOx formation. Diesel particulate filters are an integral part of any low-pressure EGR
system because they are needed to ensure that large amounts of particulate matter are not
re-circulated to the engine. In mobile source applications, NOx reductions of
approximately 40% have been reported. Figure VI-13 depicts a low-pressure EGR system
for diesel engines.61
61
“Closed Crankcase Filtration - The Next Step in Diesel Engine Emissions Reduction,” Marty
Barris, Donaldson Company, Inc. (as printed in the Summer 2000 edition of the Clean Air
Technology News, a joint publication of the Manufacturers of Emission Controls Association and
the Institute of Clean Air Companies. Washington, DC, September 2000.)
74
Figure VI-13
Exhaust Gas Recirculation
7. Other NOx Control Technologies and Strategies for Lean-Burn Engines
Lean NOx catalyst systems have also been used on lean-burn engines. Some lean NOx
catalysts rely on the injection of a small amount of diesel fuel or other reductant into the
exhaust. The fuel or other hydrocarbon reductant serves as a reducing agent for the
catalytic conversion of NOx to N2. Other systems operate passively at reduced NOx
conversion rates. The catalyst substrate is a porous material often made of zeolite. The
substrate provides microscopic sites that are fuel/hydrocarbon rich where reduction
reactions can take place. Without the added fuel and catalyst, reduction reactions that
convert NOx to N2 would not take place because of the excess oxygen present in the
exhaust. An HC to NOx ratio of up to 6:1 is needed to achieve optimal NOx reductions.
Since the fuel used to reduce NOx does not produce mechanical energy, lean NOx
catalysts typically operate with a fuel penalty of about 3%. Currently, peak NOx
conversion efficiencies are typically around 10-20%.
Two types of lean NOx catalyst formulations have emerged: a low temperature catalyst
based on platinum and a high temperature catalyst utilizing base metals, usually copper.
Each catalyst is capable of controlling NOx over a narrow temperature range. Combining
high and low temperature lean NOx catalyst systems broadens the temperature range over
which they convert NOx, making them more suitable for practical applications.
75
Engine timing retard has also been demonstrated to provide NOx reductions in the range
of 15-30% for stationary diesel IC engines.62 This is achieved by retarding the timing by
four degrees top dead center. The retarded timing provides lower combustion
temperatures which results in lower NOx emissions. However, CO and PM emissions
typically increase. This increase could be more than offset with the combined use of a
diesel particulate filter or oxidation catalyst.
NOx adsorber catalysts are currently undergoing extensive research and development in
anticipation of the new federal on-road heavy-duty diesel engine regulations scheduled to
take effect in 2007. NOx adsorbers act to store NOx emissions during lean engine
operation and release the stored NOx by periodically creating a rich exhaust environment,
either through engine operation or by injecting a reductant in the exhaust stream. When
released, the NOx is converted to N2 by a three-way catalytic reaction. Although the use
of NOx adsorber catalysts presently is not feasible on stationary diesel engines, NOx
adsorber technology may be available for use on stationary engines in the future.
8. Ultra-Low Sulfur Fuel as a Control Option for Particulate Matter
The use of low sulfur fuel can also be used as a strategy to reduce engine-out particulate
emissions from stationary diesel engines. In a recent manufacturers’ study, switching
from 368 ppm sulfur fuel to 54 ppm sulfur fuel reduced engine-out PM emissions from
0.073 g/bhp-hr to 0.063 g/bhp-hr, or by almost 14% as measured over the FTP.63
62
U.S. Environmental Protection Agency, OAQPS. EPA Air Pollution Cost Manual, Sixth Edition. EPA452-02-001, January 2002.
63
Manufacturers of Emission Controls Association, Stationary Engine Emission Control, May 2002.
76
VII. Control Technology Case Studies
To provide better information on the “real-world” cost, operation and performance of
potential control technologies for stationary diesel IC engines, ESI International prepared
case studies of six actual control technology installations. The case studies use data
obtained from engine operators to develop cost and cost-effectiveness data on various
control technology options. Since stationary IC engines equipped with emission controls
are not summarized in a national database, several means were used to identify likely
study participants. Most study participants were identified through state air pollution
control agencies. Attempts were also made to identify study participants by contacting
emission control suppliers and by searching the U.S. EPA’s RACT/BACT/LAER (RBL)
Clearinghouse. The RBL Clearinghouse contains information concerning air permits
issued by state and local air pollution control agencies in the United States. A rigorous
statistical method was not used to select a sample of case study participants because the
universe of engines equipped with emission controls is not well defined and the principal
reason for “selecting” a facility was its willingness to participate in a study.
The case studies were performed in close cooperation with the facility owners. In each
case the facility owners filled out a questionnaire that requested pertinent information.
(included as Appendix E). After information was gathered from the owners, a case study
was prepared and then sent to the facility owners for review and comment. Completed
questionnaires along with cost data, maintenance information and other documentation
were used to make calculations of emissions, cost and cost-effectiveness.
Table VII-1 summarizes the case studies developed for this report. Four of the case
studies examine the use of particulate filters, one analyzes the combined application of
particulate filters and SCR, and one involves a control installation that employed
oxidation catalysts. More detailed descriptions of each case study including specific
engine, cost and contact information follow. However, this is preceded by a description
of the methodology used to develop cost figures for each of the case studies.
77
Table VII-1
Summary of Case Study Facilities
Facility
Location
Engine
Kings County
Dept. of Public
Works
National Steel
and Shipping
Company
Pacific Bell-SBC
Kings County, Caterpillar 3516B:
CA
2848 hp @ 1800 rpm
Cummins QST30 G1: Particulate
1030 hp @ 1800 rpm Filter and
SCR
San
Caterpillar 3516:
Particulate
Francisco, CA 2841 hp @ 1800 rpm Filter
Pacific Bell-SBC San Jose, CA Cummins KTA50-64: Oxidation
2200 hp @ 1800 rpm Catalysts
Santa Clara
Santa Clara
Cummins KTA50
Particulate
County Building County, CA
G2: 2200 hp @ 1800 Filter
Operations
rpm
Sierra Nevada
Chico, CA
Caterpillar 3412:
Particulate
Brewing Co.
1109 hp @ 1800 rpm Filter
a
San Diego,
CA
Control
Technology
Particulate
Filter
Emissions
Controlled
PM, CO,
and HC
PM, NOx,
CO, and
HC
PM, CO,
and HC
PM, CO,
and HC
PM, CO,
and HC
PM, NOx,a
CO, and
HC
As discussed in the case study.
A. Cost Methodology
With the exception of a few modified assumptions, this study used cost estimating
procedures developed by EPA’s Office of Air Quality Planning and Standards (OAQPS)
and described in the OAQPS Control Cost Manual.64 Generally, these procedures allow
costs to be analyzed on an annual basis assuming equal end-of-year costs, including
direct and indirect annual costs and costs associated with recovering capital expenditures
for purchased air pollution control equipment. Annual capital recovery costs were
determined by multiplying the total capital cost of the emission control equipment by a
capital recovery factor (CRF). This factor, explained in detail in the OAQPS document,
uses assumptions regarding the annual pretax marginal rate of return on private
investment, and expected project life to create a stream of equal payments (capital
recovery costs) that recoup capital costs over the life of the project.
In following the OAQPS procedure, participants were asked to provide accurate cost
information for the major capital and annual cost categories listed below.
64
U.S. Environmental Protection Agency, OAQPS, EPA Air Pollution Cost Manual, Sixth Edition. EPA452-02-001, January 2002.
78
•
•
Capital Costs
− Purchase equipment costs
− Direct installation costs
− Indirect installation costs
− Contingencies
Annual Costs
− Direct annual costs
− Indirect annual costs
In many cases, participants provided aggregate costs (e.g. total capital costs that included
purchase equipment costs, direct installation costs, indirect costs and contingencies).
When this type of cost information was provided, no attempt was made to disaggregate
the costs into subordinate costs.
Since the following operation and maintenance costs can contribute strongly to annual
costs, an effort was made to collect detailed cost information for these items.
•
•
•
•
•
•
•
Utilities
Operating labor
Operating materials (e.g. reagent use)
Maintenance labor
Maintenance materials
Cost of an annual compliance test
Catalyst replacement
While more elaborate techniques can be used to estimate the costs of purchasing and
operating emission control equipment, the procedure developed by OAQPS provides
reasonably accurate cost data for regulatory purposes. The procedure has been used by
EPA and other organizations to evaluate the costs of many different types of control
equipment.
B. Case Study Assumptions
The following assumptions were made in all of the case studies presented:
1.
The annual pretax marginal rate of return on private investment was assumed to
be 8%. The OAQPS cost estimating procedure, last updated in 1990, uses a 10%
return on investment. More recent studies have used a rate of return of 8%.65
Because 8% better captures return on investment under current economic
conditions, this figure was used in calculating the capital recovery factor (CRF).
65
See, for example, NESCAUM’s Status Report on NOx Controls for Gas Turbines, Cement Kilns,
Industrial Boilers, Internal Combustion Engines – Technologies & Cost Effectiveness. December 2000.
79
2.
A 10-year project life was used in this collection of case studies to be consistent
with previous studies.
3.
Emissions reductions were calculated using source test data when available.
When source test data were not available, emissions reductions were based on:
•
•
•
the performance claims of the emission control manufacturers,
reductions demonstrated on similar equipment under similar operating
conditions, and
engineering judgment.
4.
When a case study participant provided aggregate capital costs for both NOx and
PM controls, costs were disaggregated by estimating the cost of the PM control
system and then subtracting PM control costs from total costs to derive an
estimate of NOx control costs. PM control costs were estimated using a cost
factor of $22 per horsepower derived from current diesel retrofits in California.
5.
When a study participant provided aggregate operating or maintenance costs for
NOx and PM controls, annual operating costs for each pollutant were
disaggregated by multiplying total annual costs by the ratio of NOx capital costs
to total capital costs and PM capital cost to total capital costs, respectively.
6.
Annual overhead costs were estimated to be 15% of annual labor costs.
C. Case Study Descriptions
1. Kings County, Department of Public Works, CA
Facility Description
Kings County is located in California’s San Joaquin Valley about halfway between Los
Angeles and San Francisco. Hanford, the county seat, is located about 210 miles
southeast of San Francisco. To save on electricity costs, the County subscribes to an
interruptible power program with the local utility. As a program participant, the County
gets a 25% lower rate on electricity but must disconnect from utility grid after being
given 30-minute notice. If the County fails to disconnect in 30 minutes, it must pay a
large penalty. During interruptions in prior years, the County would shed all load and run
essential services on small local generators. After a period of frequent interruptions
during the winter of 2000/2001, the County Council authorized the installation of a large
generator to supply backup power to most facilities located at Government Center in
Hanford. The 2 MW generator that was installed supplies backup power to essential
services including the jails, juvenile detention facilities, emergency dispatch (911) center
and the mainframe computer system. The generator also provides backup power for the
County Courts, the Department of Human Services, the Department of Health Services
and other government offices.
80
A Caterpillar 3516B engine powers the backup generator. The turbocharged and
aftercooled engine produces 2,848 hp at 1,800 rpm. To comply with air quality
regulations for diesel generators that run during voluntary power interruptions, the
County installed a diesel particulate filter system on the engine. Installation of the filter
system was a joint effort involving the County, the system supplier, and the filter
installer. The filter system supplier benefited from the project in that it gave them the
opportunity to certify their filter system with the California Air Resources Board (CARB)
under CARB’s Diesel Emission Control Strategy Verification Program. The generator
engine burns ultra-low sulfur diesel fuel, fuel that contains less than 15 ppm sulfur. The
engine operated by the Kings County Department of Public Works is summarized in
Table VII-2.
Table VII-2
Case Study 1: Engine and Emission Control Summary
Engine
Horsepower
Engine Age
Fuel
Emission Controls
Caterpillar 3516B
2,848 hp @ 1,800 rpm
1 year
Ultra low sulfur diesel fuel
Diesel particulate filter
Emissions
Engine emissions before and after installation of the particulate filters are summarized in
Table VII-3. Emissions before installation of the filter system were provided by
Caterpillar, Inc. and represent emissions for the engine operating at 50% load, the
average load the engine carries at the Government Center. Engine-out emissions for
particulate matter were also reduced by 25% as suggested by Caterpillar, Inc. to account
for the use of cleaner burning California diesel fuel. Since Kings County runs their
engine on ultra-low sulfur diesel fuel, a fuel with lower sulfur content than California
diesel fuel, actual particulate matter emissions may be even lower than those reported in
this case study. Emissions reductions were calculated from percent reduction information
obtained by the filter manufacturer. The percent reduction information was based on
actual source test data.
Table VII-3
Case Study 1: Emissions Reduction Summary
Emission Rate Before Emission Rate After
Pollutant
Controls (g/bhp-hr)
Controls (g/bhp-hr)
Carbon Monoxide
0.84
0.084
Hydrocarbons
0.33
0.033
a
Particulate Matter
0.179
0.027a
Nitrogen Oxides
6.51
NAb
a
Percent
Reduction
90
90
85
NAb
PM emission rate reduced 25% per Caterpillar, Inc. to account for lower PM emissions from CARB
diesel fuel
b
not applicable
81
The air quality permit for this engine limits its operations to 614 hours per year.
Emissions calculations were based on this maximum number of operating hours.
Emissions were estimated using a load factor of 50%. The 50% load factor was based on
actual data provided by Kings County for average weekday summer load.
Costs
The total capital cost of the filter system was estimated at $121,153 in year 2000 dollars.
This cost includes the cost of the filters and direct and indirect installation costs. Both the
particulate filter supplier and Kings County provided purchased equipment cost
information from which this total capital costs were derived.
Kings County does not expect to incur annual operation and maintenance (O&M) costs
for the filter system because the engine operates at sufficient load to generate the exhaust
gas temperatures to hot enough to oxidize accumulated particulate matter in the filter
system. To date, the filter system has been maintenance free. The County is not required
to conduct periodic compliance tests, so compliance test costs were not included in
annual costs. Since O&M costs for the filter system are expected to be negligible,
overhead – which is estimated from annual labor costs – was estimated to be negligible.
Annual costs for the filter system are summarized in Table VII-4.
82
Table VII-4
Case Study 1: Annual DPF Costs and Emissions
Costsa
Annual Costs
Direct Costs
Labor and Materials
Compliance Test
Total Direct Costs
Indirect Costs
Overhead
Capital Recovery Cost
Total Indirect Costs
Total Annual Costs
$0
$0
$0
$0
$18,055
$18,055
$18,055
Annual Emissions
PM Emissions Before Installation of Controls
PM Emissions After Installation of Controls
Annual Tons of PM Removed
Percent Reduction
a
Tons per Year
0.173
0.026
0.147
85%
CO Emissions Before Installation of Controls
CO Emissions After Installation of Controls
Annual Tons of CO Removed
0.810
0.081
0.729
Percent Reduction
90%
HC Emissions Before Installation of Controls
HC Emissions After Installation of Controls
Annual Tons of HC Removed
0.318
0.033
0.032
Percent Reduction
90%
Total Annual Tons of Pollution Reduced
0.908
year 2000 dollars
Facility and Contact Information
Kings County Department of Public Works
1400 W. Lacey Boulevard
Hanford, CA 93230
Mr. Harry W. Verheul
(559) 582-3211 ext. 2698
83
2. National Steel and Shipbuilding Company (NASSCO)
Facility Description
NASSCO is the largest shipyard engaged in new ship construction on the West Coast.
The shipyard, which is located on San Diego Bay, builds commercial ships including oil
tankers, ferries, container ships and research vessels. It also builds U.S. Navy support
ships.
In 2001, the shipyard purchased a Cummins QST30-G1 diesel engine and generator to
power Crane Number 16, a 300-ton gantry crane. The engine, rated at 1,135 hp at 1,800
rpm, is turbocharged and aftercooled. To meet air quality requirements, the engine was
equipped with Best Available Control Technology (BACT). The air pollution control
equipment installed on the engine included a diesel particulate filter to control particulate
matter and a selective catalytic reduction (SCR) unit to reduce NOx emissions. The
reagent used in the unit is a 40% aqueous solution of urea. The reagent is consumed at a
rate of about 2.7 pints per hour (0.34 gallons per hour). The San Diego County Air
Pollution Control District limits ammonia emissions (ammonia slip) to less than 10 ppm.
The NASSCO SCR unit operates within that limit.
To keep the exhaust gas temperatures hot enough for proper operation of the particulate
filter and SCR system, exhaust gas heaters are installed ahead of the emission control
system. Exhaust gas temperatures are maintained above 715º F to oxidize accumulated
particulate matter in the filter system. The SCR system needs temperatures above 570º F
to maintain NOx conversion efficiency.
NASSCO operates Crane Number 16 on California #2 diesel fuel. Table VII-5
summarizes the equipment installed at NASSCO.
Table VII-5
Case Study 2: Engine and Emission Control Summary
Engines
Horsepower
Engine Age
Fuel
Emission Controls
Cummins QST30-G1
1,030 hp @ 1,800 rpm
2 years
CARB #2 diesel fuel
Selective catalytic reduction (SCR) & Diesel
particulate filter (DPF) system
Emissions
Engine emissions before and after installation of emission controls are summarized in
Table VII-6. The emissions rates for the engine before the installation of emission
controls are from Cummins, Inc. The NASSCO crane engine usually operates at low
load, or approximately 11-15% capacity. Calculated loads for the case study were 12.8%.
Since emissions data were not available for the Cummins QST30 engine at low load,
Cummins provided emissions data for a similar 1,525 hp engine operating at 10% load.
84
Emissions rates at the 10% power point for the 1,525 hp engine approximate emissions
rates for the 1,135 hp engine operating at 13.5% load, roughly the load observed at
NASSCO.
Emissions reductions are based on the performance claims of the emission control
manufacturer and reflect typical values observed in similar applications. The emission
control installer estimates the diesel particulate filter achieves an 85% reduction of
particulate matter and the SCR unit achieves a 90% reduction of NOx. Source tests were
conducted on the crane by World Environmental, Inc. on March 14, 2002. Preliminary
source test data confirm the SCR system achieves a >90% reduction of NOx. However,
PM emissions appeared to increase. This is still under investigation and may be due to a
sampling anomaly. Because the control efficiency of diesel particulate filters is well
known, an 85% reduction is assumed.
Table VII-6
Case Study 2: Emissions Reduction Summary
Pollutant
a
Carbon Monoxide
Hydrocarbons
Particulate Matter
Nitrogen Oxides
Emission Rate
Before Controls
(g/bhp-hr)
0.2
1.1
0.03a
8.1b
Emission Rate After
Controls (g/bhp-hr)
Percent
Reduction
0.85
0.10
0.045a
0.69b
85
85
85
93
PM emission rates reduced 25% per Caterpillar, Inc. to account for lower PM emissions from
California diesel fuel
b
NOx emission rates reduced 7% per Caterpillar, Inc. to account for lower NOx emissions from
California diesel fuel
NASSCO operates Crane Number 16 two eight-hour shifts per day, six days a week, 50
weeks per year. Thus, the crane operates about 4,800 hours per year. An engine load
factor of 0.12 was calculated from fuel consumption data provided by NASSCO. The
load factor appears reasonable given that the crane spends large amounts of time at idle.
The 0.12 load factor was used in all annual emissions calculations.
Costs
NASSCO estimated the capital cost of the SCR and DPF control system for the engine as
$289,900 in year 2000 dollars. This cost included purchase equipment costs and direct
and indirect installation costs. The capital cost for the filter system, including the exhaust
gas heaters, was approximately $111,020. The capital cost of the SCR system was
$178,880.
Annual costs for the diesel particulate filter system included maintenance labor, prorated
costs for a semiannual compliance test, overhead and recovery of capital. No filter
replacement costs were assumed for the ten-year life of the filter system. The most
significant annual cost associated with the filter system was the cost of recovering capital.
Cost information for the filter system is summarized in Table VII-7.
85
Table VII-7
Case Study 2: Annual DPF Costs and Emissions
Costsa
Annual Costs
Direct Costs
Labor
Compliance Test
Total Direct Costs
Indirect Costs
Overhead
Capital Recovery Cost
Total Indirect Costs
Total Annual Costs
$2,930
$1,862
$4,791
$439
$16,545
$16,985
$21,776
Annual Emissions
PM Emissions Before Installation of Controls
PM Emissions After Installation of Controls
Annual Tons of PM Removed
a
Tons per Year
0.023
0.003
0.020
Percent Reduction
85%
CO Emissions Before Installation of Controls
CO Emissions After Installation of Controls
Annual Tons of CO Removed
0.154
0.015
0.139
Percent Reduction
90%
HC Emissions Before Installation of Controls
HC Emissions After Installation of Controls
Annual Tons of HC Removed
0.848
0.085
0.763
Percent Reduction
90%
Total Annual Tons of Pollution Reduced
0.922
year 2000 dollars
Annual O&M costs for the SCR system include the cost of urea reagent, maintenance
labor, prorated costs for a semiannual compliance test and the levelized cost of replacing
the SCR catalyst. Urea is purchased at $2.71 per gallon and supplied to the SCR system
at a dosing rate of 0.34 gallons per hour. The SCR supplier has recommended that
NASSCO follow a phased replacement schedule for the SCR catalyst replacing one of the
three SCR catalyst layers every 20,000 operating hours. Each layer costs $2,600. Under
NASSCO’s current operating scenario, the levelized cost of SCR catalyst replacement is
$624 per year. Cost information for the SCR system is summarized in Table VII-8.
86
Table VII-8
Case Study 2: Annual SCR Costs and Emissions
Costsa
Annual Costs
Direct Costs
Reagent
Labor
Compliance Test
Catalyst Replacement Costs
Total Direct Costs
Indirect Costs
Overhead
Capital Recovery Cost
Total Indirect Costs
Total Annual Costs
$4,390
$4,720
$3,000
$624
$12,734
$4,720
$26,658
$31,379
$44,113
Emissions
NOx Emissions Before Installation of Controls
NOx Emissions After Installation of Controls
Annual Tons of NOx Removed
a
Tons per Year
6.24
0.67
5.57
Percent Reduction
90%
year 2000 dollars
Facility and Contact Information
National Steel & Shipbuilding Company
Harbor Drive & 28th Street
P.O. Box 85278
San Diego, CA 92186-5278
Mr. Dina Ahmed
(619) 544-8764
3. Pacific Bell-SBC Telecommunications Facility, San Francisco, CA
Facility Description
In 1994, Pacific Bell-SBC installed two emergency generators at their high-rise
telecommunications facility at 611 Folsom Way, San Francisco, CA. Originally at this
location, Pacific Bell operated a turbine to provide emergency backup power. To
accommodate growth, Pacific Bell replaced the turbine with two Caterpillar 3516 engines
and their associated generators. Each engine produces 2,841 horsepower at 1,800 rpm.
Operation of the earlier turbine had resulted in several nuisance complaints concerning
smoke and odors. To avoid future complaints, Pacific Bell installed diesel particulate
87
filter systems on the new Caterpillar engines. The installed filter systems are capable of
reducing particulate matter emissions by 85% and carbon monoxide and hydrocarbon
emissions by 90%. The engines burn regular #2 diesel fuel. Equipment installed at Pacific
Bell’s telecommunications facility at 611 Folsom Way is summarized in Table VII-9.
Table VII-9
Case Study 3: Engine and Emission Control Summary
Engine
Horsepower
Engine Age
Fuel
Emission Controls
Caterpillar 3516
2,841 hp @ 1,800 rpm
8 years
#2 diesel fuel
Diesel particulate filter
Emissions
Engine emissions before and after installation of the particulate filters are summarized in
Table VII-10. Emissions reductions are based on the emissions reduction claims of the
control system manufacturer.
Table VII-10
Case Study 3: Emissions Reduction Summary
Pollutant
a
Carbon Monoxide
Hydrocarbons
Particulate Matter
Nitrogen Oxides
Emission Rate
Before Controls
(g/bhp-hr)
1.17
0.50
0.239
17.04
Emission Rate After
Controls (g/bhp-hr)
Percent
Reduction
0.12
0.05
0.036
NAa
90
90
85
NAa
not applicable
Electricity service is quite reliable in the San Francisco area and as a result, the engines
are operated only about 20 hours per year. Pacific Bell runs the engines about one hour
every month to ensure they will be able to carry building load in an emergency. Once a
year, the engines are run for longer periods to test transfer switchgear and other
equipment.
Actual load data were not available to estimate a load factor for the engine. A load factor
of 0.15, a typical load factor for maintenance-type operation, was used to make emissions
estimates.
Costs
The total capital cost of a filter system for one of the engines was $95,860 in year 2000
dollars. This cost included the cost of the filters, electric heaters, structural supports,
88
exhaust system piping and insulation, controls and instrumentation, and other direct and
indirect installation costs.
Pacific Bell estimates the annual cost of operating and maintaining a single filter system
is about $5,000. Pacific Bell is not required to conduct periodic compliance tests, so no
compliance test costs were included in annual costs. Overhead, which is estimated from
annual labor costs, was estimated to be about $750 per year. Costs for a single filter
system are summarized in Table VII-11.
Table VII-11
Case Study 3: Annual DPF Costs and Emissions
Costsa
Annual Costs
Direct Costs
Labor and Materials
Compliance Test
Total Direct Costs
Indirect Costs
Overhead
Capital Recovery Cost
Total Indirect Costs
Total Annual Costs
$5,000
$0
$5,000
$750
$14,286
$15,036
$20,036
Annual Emissions
PM Emissions Before Installation of Controls
PM Emissions After Installation of Controls
Annual Tons of PM Removed
Percent Reduction
85%
CO Emissions Before Installation of Controls
CO Emissions After Installation of Controls
Annual Tons of CO Removed
0.009
0.001
0.008
Percent Reduction
90%
HC Emissions Before Installation of Controls
HC Emissions After Installation of Controls
Annual Tons of HC Removed
Percent Reduction
a
Tons per Year
0.002
0.000
0.002
0.004
0.0004
0.0036
90%
Total Annual Tons of Pollution Reduced
year 2000 dollars
89
0.0014
Facility and Contact Information
Pacific Bell-SBC
95 Almaden Street
Room 316
San Jose, CA 95113
Ms. Lynn Bowers
(408) 491-2402
4. Pacific Bell-SBC Telecommunications Facility, San Jose, CA
Facility Description
In 2000, Pacific Bell-SBC installed an emergency generator at their telecommunications
facility in a rural area near San Jose, CA. The telecommunications center, located at 6245
Dial Way, serves as a dial tone, high-speed data and interconnection facility. The engine
installed at the site is a Cummins KTA50-G9 producing 2,220 horsepower at 1,800 rpm.
It is turbocharged and aftercooled. When the engine was installed, Pacific Bell equipped
the engine with a diesel oxidation catalyst to control exhaust odors. The exhaust stack for
the engine is near air intakes for a neighboring building and there was concern that
exhaust could find its way into the neighboring building. The engine operates on a
mixture of #1 and #2 diesel fuel. Equipment installed at Pacific Bell’s
telecommunications facility is summarized in Table VII-12.
Table VII-12
Case Study 4: Engine and Emission Control Summary
Engine
Horsepower
Engine Age
Fuel
Emission Controls
Cummins KTA50-G9
2,220 hp @ 1,800 rpm
2 years
#1 and #2 diesel fuel
Diesel Oxidation Catalyst
Emissions
Engine emissions before and after installation of the oxidation catalyst are summarized in
Table VII-13. Emissions reductions are based on the emissions reduction claims of the
control device manufacturer.
90
Table VII-13
Case Study 4: Emissions Reduction Summary
Emission Rate
Emission Rate After
Pollutant
Before Controls
Controls (g/bhp-hr)
(g/bhp-hr)
Carbon Monoxide
8.5
0.22
Hydrocarbons
1.0
0.01
Particulate Matter
0.4
0.01
Nitrogen Oxides
6.9
NAa
a
Percent
Reduction
90
90
25
NAa
not applicable
The emergency backup generator is only operated about 20 hours per year. Pacific Bell
runs the engine about one hour every month to exercise it. Once a year, the engine is run
for a longer period to test transfer switchgear and other equipment. The electrical grid is
stable in this area of California and the engine is rarely used to generate electricity in a
power outage.
Actual load data were not available to estimate a load factor for the engine. A load factor
of 0.15, a typical load factor for maintenance-type operation, was used to make emissions
estimates.
Costs
The total capital cost of the diesel oxidation catalyst in year 2000 dollars was $24,895.
This cost includes the cost of the emission control device, $12,950 in year 2000 dollars,
plus sales taxes, freight, direct and indirect installation costs.
Annual costs for the catalyst were minimal involving only the recovery of capital. Since
the engine operates infrequently, labor costs were estimated to be negligible. Pacific Bell
is not required to conduct periodic compliance tests, so no compliance test costs were
included in annual costs. Overhead, which is estimated from annual labor costs, is also
negligible in this case study. Cost information for the catalyst is summarized in Table
VII-14.
91
Table VII-14
Case Study 4: Annual DOC Costs and Emissions
Costsa
Annual Costs
Direct Costs
Labor
Compliance Test
Total Direct Costs
Indirect Costs
Overhead
Capital Recovery Cost
Total Indirect Costs
Total Annual Costs
$0
$0
$0
$0
$3,710
$3,710
$3,710
Annual Emissions
PM Emissions Before Installation of Controls
PM Emissions After Installation of Controls
Annual Tons of PM Removed
a
Tons per Year
0.003
0.002
0.001
Percent Reduction
25%
CO Emissions Before Installation of Controls
CO Emissions After Installation of Controls
Annual Tons of CO Removed
0.062
0.006
0.056
Percent Reduction
90%
HC Emissions Before Installation of Controls
HC Emissions After Installation of Controls
Annual Tons of HC Removed
0.007
0.001
0.007
Percent Reduction
90%
Total Annual Tons of Pollution Reduced
0.064
year 2000 dollars
Facility and Contact Information
Pacific Bell-SBC
95 Almaden Street
Room 316
San Jose, CA 95113
Ms. Lynn Bowers
(408) 491-2402
92
5. Santa Clara County Building Operations
Facility Description
Santa Clara County operates a back-up generator at 1555 Berger Drive in San Jose, CA.
The generator is located in the basement of Building 2 of the County Service Center. The
generator provides emergency back-up power to three buildings in the Service Center.
These buildings house the Departments of Agriculture, Information Services and
Revenue, the General Services Administration, the Registrar of Voters, the District
Attorney’s Crime Lab, Telephone Services, Property Management, Purchasing and other
government offices.
In 1998, Santa Clara County purchased the Cummins KTTA50-G2 diesel engine and
generator to provide emergency power for the County Service Center. The engine is rated
at 2,220 hp at 1,800 rpm. It is turbocharged and aftercooled. To avoid complaints
concerning smoke and odors, and to be a “good neighbor” to nearby building occupants,
Santa Clara County installed two catalyzed diesel particulate filters on the engine exhaust
system at the time the engine was installed. The filters significantly reduce CO, HC and
PM emissions. They were not installed to meet the requirement of an air quality permit.
The engine operates on CARB Diesel Fuel #2. Equipment installed at Santa Clara County
is summarized in Table VII-15.
Table VII-15
Case Study 5: Engine and Emission Control Summary
Engine
Horsepower
Engine Age
Fuel
Emission Controls
Cummins KTTA50-G2
2,220 hp @ 1,800 rpm
4 years
CARB #2 Diesel Fuel
Diesel Particulate Filter (DPF) system
Emissions
Engine emissions before and after installation of the filter system on the engine are
summarized in Table VII-16. Emissions reductions are based on preliminary source test
results provided by the emission control manufacturer for a similar emission control
system installed at another facility. The engine and control system at Santa Clara
County’s Service Center was emission tested by the California Air Resources Board
(CARB), however the test results were not available at the time this case study was
prepared.
93
Table VII-16
Case Study 5: Emissions Reduction Summary
Pollutant
Carbon Monoxide
Hydrocarbons
Particulate Matter
Nitrogen Oxides
a
Emission Rate
Before Controls
(g/bhp-hr)
2.17
0.14
0.083a
12.8
Emission Rate After
Controls (g/bhp-hr)
0.22
0.01
0.012a
NAb
Percent
Reduction
90
90
85
NAb
PM emission rate reduced 25% per Caterpillar, Inc. to account for lower PM emissions from CARB
diesel fuel.
b
not applicable
The number of hours the engine has been operated per year has varied since its
installation. Before the engine was issued an air quality permit, the engine operated
hundreds of hours per year when the local utility, Pacific Gas and Electric, instituted
“load curtailment.” During load curtailment, periods of high electricity demand and/or
limited electricity supply, PG&E would ask local generators to come on-line to reduce
load on the grid. Now, the air quality permit limits the County’s generator operations to
only 100 hours per year over actual emergency operations. Emergency operations are
minimal involving only one or two events per year of about two to three hours per event.
Santa Clara County estimates the engine now operates only about 70 hours per year.
Actual load data were not available to estimate a load factor for the engine. Given the
lack of actual data, a load factor of 0.15 was used to make emissions estimates. This load
factor represents the typical load standby engines carry when they are exercised monthly
or weekly under minimal load.
Costs
The total capital cost of the DPF system, $45,834 in year 2000 dollars, includes the actual
cost of the DPF system, $24,000, plus additional amounts for installation and other costs.
Annual costs for the filter system were minimal involving only the recovery of capital.
Since the engine operates infrequently, labor costs were estimated to be negligible. Santa
Clara County is not required to conduct periodic compliance tests, so no compliance test
costs were included in annual costs. Overhead, which is estimated from annual labor
costs, is also negligible in this case study. Cost information for a filter system is
summarized in Table VII-17.
94
Table VII-17
Case Study 5: Annual DPF Costs and Emissions
Costsa
Annual Costs
Direct Costs
Labor
Compliance Test
Total Direct Costs
Indirect Costs
Overhead
Capital Recovery Cost
Total Indirect Costs
Total Annual Costs
$0
$0
$0
$0
$6,831
$6,831
$6,831
Annual Emissions
PM Emissions Before Installation of Controls
PM Emissions After Installation of Controls
Annual Tons of PM Removed
Percent Reduction
85%
CO Emissions Before Installation of Controls
CO Emissions After Installation of Controls
Annual Tons of CO Removed
0.056
0.006
0.005
Percent Reduction
90%
HC Emissions Before Installation of Controls
HC Emissions After Installation of Controls
Annual Tons of HC Removed
Percent Reduction
a
Tons per Year
0.003
0.001
0.002
0.004
0.0004
0.0036
90%
Total Annual Tons of Pollution Reduced
year 2000 dollars
Facility and Contact Information
Santa Clara County
Building Operations – GSA
1555 Berger Drive
San Jose, CA 95112
Ms. Alana Crary
(408) 299-4181 ext. 2156
95
0.0536
6. Sierra Nevada Brewing Company, Chico, CA
Facility Description
Sierra Nevada Brewing Company (SNBC) is a brewery that produces ales, stouts and
beers in Chico, CA. In 1997 and 1999, the brewery purchased two 750 kW generators to
provide emergency backup power for brewery operations. Both generators are powered
by Caterpillar 3412 diesel engines. Each turbocharged engine produces 1,109 hp at 1,800
rpm. To meet air quality requirements, SNBC installed diesel particulate filters on the
engines in 1999 and 2000. Two catalyzed diesel particulate filters are installed in parallel
on each engine. The engines run on CARB diesel fuel. The engines operated by SNBC
are summarized in Table VII-18.
Table VII-18
Case Study 6: Engine and Emission Control Summary
Engines
Horsepower
Engine Ages
Fuel
Emission Controls
Caterpillar 3412
1,109 hp @ 1,800 rpm
3 years and 5 years
CARB diesel fuel
Diesel particulate filters
Emissions
Engine emissions before and after installation of the particulate filters are summarized in
Table VII-19. Emissions before installation of the filter system were provided by
Caterpillar, Inc. and are emissions for the engine operating at 75% load. SNBC typically
operates their engines at about 80% load. Emissions data at 80% load were not available
from Caterpillar, Inc. Uncontrolled engine emissions for NOx and PM were reduced by
7% and 25%, respectively, as suggested by Caterpillar, Inc. to account for the use of
cleaner burning CARB diesel fuel. Occasionally, SNBC runs their engines on a diesel
fuel that contains less sulfur than CARB diesel fuel. When this lower sulfur fuel is used,
actual PM emissions may be even lower than those reported in this case study.
CARB performed source tests on SNBC’s engines on March 20-24, 2000. To the extent
data were available, Caterpillar engine emissions data and CARB source test data were
used to calculate emissions reductions. Average source test emissions for NOx, converted
to grams per brake horsepower, were subtracted from Caterpillar engine-out emissions to
calculate the 12.3% drop in NOx emissions reported below. The CARB source test report
concluded PM test results should be viewed as only “qualitative in nature” because of
testing anomalies. PM emissions reductions calculated from CARB source test data agree
well, however, with reductions observed in other similar applications of diesel particulate
filter technology. Average source test emissions for PM, converted to grams per brake
horsepower, subtracted from Caterpillar engine-out emissions resulted in an emissions
reduction of 85%. Reductions of this magnitude are often observed when diesel
particulate filters are installed on stationary diesel engines.
96
Table VII-19
Case Study 6: Emissions Reduction Summary
Pollutant
Carbon Monoxide
Hydrocarbons
Particulate Matter
Nitrogen Oxides
a
Emission Rate
Before Controls
(g/bhp-hr)
0.84
0.33
0.164a
6.69b
Emission Rate After
Controls (g/bhp-hr)
Percent
Reduction
0.084
0.033
0.025a
5.85
90
90
85
12
PM emission rate reduced 25% per Caterpillar, Inc. to account for lower PM emissions from CARB
diesel fuel.
b
NOx emission rate reduced 7% per Caterpillar, Inc. to account for lower NOx emissions from CARB
diesel fuel.
SNBC’s air quality permit limits engine operations and each engine operates only 100 to
150 hours per year. Most operating hours are devoted to exercising the engines to ensure
they are ready to come on-line in an emergency. Emission calculations were based on one
engine operating 150 hours per year. Emissions were estimated using a load factor of 80
percent. SNBC exercises their engines under load to achieve the engine exhaust
temperatures needed to oxidize the accumulated particulate matter on their diesel
particulate filters.
Costs
The total capital cost for a single filter system was estimated to be $40,655 in year 2000
dollars. This cost includes the cost of the filters and direct and indirect installation costs.
The cost of the initial compliance test for an engine, $1,500, was included in indirect
costs.
SNBC does not expect to incur annual operation and maintenance (O&M) costs for the
filter systems because the generator engines are exercised at sufficient load to generate
the exhaust gas temperatures hot enough to oxidize accumulated particulate matter in the
filter system. To date, the filter system has been maintenance-free. SNBC is not required
to perform periodic compliance tests so compliance test costs were not included in annual
costs. Because O&M costs for the filter systems are negligible, overhead, which is
estimated from annual labor costs, was assumed to be negligible. Annual costs for the
filter system are summarized in Table VII-20.
97
Table VII-20
Case Study 6: Annual DPF Costs and Emissions
Costsa
Annual Costs
Direct Costs
Labor and Materials
Compliance Test
Total Direct Costs
Indirect Costs
Overhead
Capital Recovery Cost
Total Indirect Costs
Total Annual Costs
$0
$0
$0
$0
$6,059
$6,059
$6,059
Annual Emissions
NOx Emissions Before Installation of Controls
NOx Emissions After Installation of Controls
Annual Tons of NOx Removed
a
Tons per Year
0.981
0.860
0.121
Percent Reduction
12%
PM Emissions Before Installation of Controls
PM Emissions After Installation of Controls
Annual Tons of PM Removed
0.024
0.004
0.020
Percent Reduction
85%
CO Emissions Before Installation of Controls
CO Emissions After Installation of Controls
Annual Tons of CO Removed
0.173
0.021
0.161
Percent Reduction
93%
HC Emissions Before Installation of Controls
HC Emissions After Installation of Controls
Annual Tons of HC Removed
0.026
0.003
0.024
Percent Reduction
90%
Total Annual Tons of Pollution Reduced
0.335
year 2000 dollars
98
Facility and Contact Information
Sierra Nevada Brewing Company
1075 East 20th Street
Chico, CA 95928
Mr. Ken Grossman
(530) 893-3520
D. Cost-Effectiveness Analysis
Based on the case studies described above, the capital cost of a filter system ranges from
about $45,000 to about $120,000. The sole diesel oxidation catalyst installation, by
contrast, had a more modest capital cost of $25,000. The capital cost of the SCR
installation was estimated to be just under $180,000.
The cost effectiveness of a control technology is typically measured in terms of control
costs incurred for a given amount of pollution removed (e.g. $/ton), where control costs
include the annual levelized capital costs of the technology as well as fixed and variable
operating and maintenance costs. Because this metric takes into account the amount of
pollution removed, it is dependent on engine operation (fuel consumption and load) and
hours of operation, as well as on the control technology used and the levels of pollution
reduction achieved. If an engine runs for only a few tens or hundreds of hours in a year, a
given control technology will likely remove only relatively small amounts of emissions
(say, compared to a central-station power plant) for a given capital cost.
To illustrate this relationship, data obtained from the case studies for both control systems
costs and emission control performance were used to determine cost-effectiveness per ton
of pollutant reduced compared to annual operating hours and type of operation as
reflected by different load factors. In calculating cost-effectiveness, all pollutant
reductions being achieved at the case study facilities were included. In other words, the
calculation includes PM, CO and HC reductions for the Kings County, Santa Clara
County and Pacific Bell (San Francisco and San Jose) installations and PM, NOx, CO
and HC for the NASSCO and Sierra Nevada installations.
To calculate engine-out emissions per year for each case study, the following equation
was used:
Mass Emission Rate x Load Factor x Horsepower x Operating Hours per Year = Mass Emissions per Year
An example calculation is shown below:
6.9 g/bhp-hr x 0.12 load factor x 1,030 bhp x 4,800 hr/yr = 4.1x106 g/yr
where:
g/bhp-hr = grams per brake horsepower hour
bhp = brake horsepower
hr = hours
yr = year
99
Converting grams to tons yields annual emissions in tons per year:
(4.1x106 g/yr) / (453.59 g/lb x 2,000 lb/ton) = 4.5 tons/yr
Figure VII-1 shows how cost-effectiveness varies depending on operating hours for each
of the case studies analyzed. At 500 annual operating hours, for example, costeffectiveness ranges from approximately $2,000/ton in the case of the Pacific Bell
installation in San Jose to $90,000/ton at National Steel and Shipbuilding Company
(NASSCO), with an average cost-effectiveness of $36,000 per ton of pollution controlled.
If the engines are assumed to operate for 2000 hours per year, these values change to a
low of less than $1,000/ton to a maximum of $23,000/ton. Average cost-effectiveness at
2000 hours of potential operation annually is $8,500 per ton of pollution reduced, based
on information provided by the case studies.
Figure VII-1
Cost-Effectiveness Versus Annual Operating Hours
90
100
90
65
70
60
46
500 Hours
50
2000 Hours
16
17
23
5
3
2
4
9
2
1
1
10
5
20
19
30
1000 Hours
32
40
10
$ "000/ton
80
0
s
ng
Ki
un
Co
ty
J
a
F
ty
CO
ad
-S
-S
un
S
v
l
l
o
l
l
e
S
C
N
Be
Be
a
NA
ra
ic
ic
ar
f
f
l
i
i
er
c
i
c
C
S
a
Pa
Pa
nt
a
S
Of the case studies presented in this report, however, only two involve engines that are
operated for other than emergency or back-up purposes – Kings County and NASSCO.
The others are operated predominantly to insure they will be available if needed and
hence operate even less frequently than the 500 hour minimum shown in Figure VII-1.
The installation at Kings County is operated to provide electricity with a 50% load factor
for up to 614 hours per year. At NASSCO, the engine is used to power a 300-ton gantry
crane for 4,800 hours per year at a load factor of only 12% because the engine spends a
great deal of time at idle. The Kings County installation is equipped with a diesel
particulate filter system for the control of PM, CO, and HC emissions. The NASSCO
facility has a combined diesel particulate filter/SCR system for the control of PM, CO,
HC, and NOx. The cost-effectiveness of controlling emissions from these two
installations using their actual operating hours is shown in Figure VII-2.
100
Figure VII-2
Control Cost-Effectiveness for Actual Operation at Kings County and NASSCO
.5
18
15
16
14
9.
5
$ '000/ton
12
10
8
6
4
2
0
s
ng
Ki
ty
un
o
C
O
SC
S
NA
As shown, the cost-effectiveness of emission control technologies for these two
operations ranges from approximately $9,500 to $15,500 per ton of pollution reduced
annually.
Cost-effectiveness is far less favorable for the other four case studies if actual operating
hours are assumed. At the Pacific Bell facilities, the subject engines are operated at low
load for only 20 hours per year. The Santa Clara County installation is operated for 70
hours annually with over 90% of the operation for testing and routine maintenance.
Similarly, the Sierra Nevada Brewing Company operates its engines at 80% load
predominantly for testing and maintenance for up to 150 hours per year. As a result, costs
per ton of pollution removed are at least an order of magnitude higher for the particulate
filter control systems installed at the Pacific Bell, Santa Clara County, and Sierra Nevada
Brewing Company facilities compared to the Kings County and NASSCO installations, if
actual operating hours are assumed. Costs per ton are somewhat lower, at $58,000/ton,
for the oxidation catalyst system installed at the Pacific Bell San Jose facility, even taking
into account its similarly low annual hours of operation.
Finally, the case studies indicate that cost-effectiveness also varies with the size of the
engine involved. For example, they suggest that current capital costs for retrofitting an
existing diesel engine with a particulate filter system are in the range of $20 to $45 per
101
horsepower. Thus, the capital cost of retrofitting a midsize, 1000 hp diesel engine
(equivalent to 750 kW), would be in the range of $20,000 to $45,000. It should be
emphasized that the relatively high capital costs associated with these control systems can
be expected to decline substantially with manufacturing economies of scale and as more
engines are retrofitted, creating a more competitive market for vendors.
102
VIII. Conclusions and Policy Recommendations
As noted in the Introduction and in subsequent chapters, two broad questions and
concerns motivated this study. The first was a recognition that environmental regulators
lack reliable and accurate information on the current population and emissions
characteristics of stationary diesel engines in the Northeast. A second, closely related
concern was the lack of information on how rapidly developing demand response and
real-time pricing programs – coupled with metering, interconnection and other
technology advances – might impact air emissions associated with diesel engines, the
most common form of distributed generation capacity available at the present time.
The findings presented elsewhere in this report suggest a number of things about the
existing stationary diesel generator population in the Northeast. First, while it is difficult
to develop reliable population and size distribution estimates for all such engines, the
total number of units distributed throughout the eight-state NESCAUM region is
significant and likely ranges from well over 12,000 units to as many as 35,000 units.
Diesel internal combustion (IC) engines clearly account for the vast majority of stationary
distributed generators installed at commercial and industrial facilities in the Northeast.
Within the diesel IC engine population, most (80%) are intended exclusively or primarily
for emergency use. Finally, a comparison of the inventory estimates and telephone
surveys conducted for this study suggests that state agencies lack permit records or other
unit-specific information on large numbers of individual engines in the region.
Available information further suggests that the impact of formal demand response
programs sponsored by New York and New England Independent System Operators
(ISOs) to date has been modest, both in terms of the total amount of customer-based
electricity generated and in terms of the pollution added to state and regional
inventories.66 In fact, the telephone surveys conducted in New York City and Fairfield
County, Connecticut, suggest that current levels of diesel generation outside formal
demand response programs are much more significant. Presumably, most of that
generation is being provided by units permitted to operate for peak-shaving and baseload
– rather than strictly emergency – purposes. At present, emergency generators are
generally precluded by state regulation from operating under all but a very limited set of
conditions. The potential for an increase in emissions from the current base of distributed
diesel generators is therefore largely driven by the following three factors:
•
•
The potential for increased capacity utilization of existing baseload or peakshaving units;
The potential for increased eligibility of – and use of – emergency engines in
emergency demand response programs, which allow for operation before an
actual outage occurs; and
66
Of course, as discussed in Chapter V, the health risks and environmental impacts of diesel generators are
perhaps more appropriately viewed from a local impacts standpoint – taking into account the unique
characteristics of diesel exhaust – than from the more traditional emissions inventory/attainment planning
standpoint in which cumulative emissions from multiple sources are often aggregated over larger areas.
103
•
The potential for increasing numbers of generators to operate to the limits of –
and perhaps in some cases even beyond – applicable air permit restrictions if
changing market signals or incentive programs provide stronger economic reasons
to do so.
At least in New England, the recent addition of substantial new central-station combinedcycle gas turbine capacity is likely to substantially ameliorate the first two, if not all, of
these potential drivers in the near term by making capacity shortfalls and high prices less
common. However, as regional demand continues to grow and investment in new centralstation capacity slows down due to a shortage of capital and loss of investor confidence
in the energy sector more broadly, shortage situations and high prices may begin to reemerge, especially in transmission-constrained load pockets. It is therefore important that
air quality management officials respond now by updating permitting and other
regulatory requirements and by ensuring that mechanisms are in place to collect the data
needed to assess and manage potential impacts in the future. In updating regulatory
requirements, states should consider the importance of limiting harmful impacts from the
existing base of stationary diesel generators, while also (ideally) promoting the transition
to a new generation of cleaner alternatives.
This chapter reviews a number of policy initiatives currently being undertaken by states
and other organizations to update existing regulatory requirements for distributed
generators and to institute better data collection practices, particularly in cooperation with
northeastern ISOs. The state summaries are followed by a brief discussion of some of the
control technology options available to diesel generators and their costs. Finally, the
chapter closes with a discussion of a number of recommendations aimed at promoting
regional policies for the use of diesel engines and other distributed generation
technologies that are protective of the environment and at the same time, provide fuel
diversity and supply reliability.
A. Emissions Standards and Model Rules for Distributed Generators
Model rules for distributed generation have been proposed by the Northeast Ozone
Transport Commission (OTC), a regional organization created under the Clean Air Act
Amendments of 1990 to address the regional transport of ozone and its precursors in the
Northeast and Mid-Atlantic states, and by the Regulatory Assistance Project (RAP), a
non-profit organization that aims to assist state and federal regulators in addressing a
wide range of electric sector issues. Both organizations solicited input from numerous
stakeholders in developing their model rules. The OTC model rule recommends fuelspecific emissions standards (for NOx only) for all non-emergency natural gas and diesel
engines.67 In addition, the OTC model rule recommends a number of requirements for
emergency engines, including requirements to (1) set and maintain engine ignition and
injection timing at specified levels; (2) inspect and adjust timing every 3 years; (3) avoid
operation on days with high ozone levels; and (4) promote record-keeping on engine
operation, including the installation of meters that can record monthly hours of operation.
67
As an alternative, the OTC model rule also allows for requirements to be met by achieving enginespecific percentage reductions from current baseline NOx emissions or by purchasing NOx allowances.
104
In contrast to the OTC model rule, the RAP model rule is aimed at new engines only and
recommends the phased introduction of progressively more stringent output-based
emissions standards for several pollutants (including particulate matter, carbon monoxide
and nitrogen oxides, as well as carbon dioxide). The emissions standards proposed in the
RAP model rule are independent of fuel type. In addition, new regulations governing
emissions from distributed generators have recently been adopted in California and
Texas. All of these efforts provide a potential source of guidance for Northeast states
looking to update their regulatory requirements for distributed generators. Table VIII-1
summarizes the specific emissions standards proposed by the OTC and RAP model rules
as well as emissions standards adopted in California and Texas.
Table VIII-1
Summary of Recommended/Adopted Distributed Generation Emissions Standards
NOx (Ozone
Attainment Areas)
Regulation
Emission Limits (lb/MWh)
NOx (Ozone NonCO
VOCs
Attainment Areas)
PM
CO2
a
OTC Model Rule - new and in use engines (emissions factors converted from g/bhp-hr)
Natural Gas (except emergency)
4.4
4.4
Diesel (except emergency)
6.8
6.8
RAP Model Rule - new engines
After January 1, 2004
4
0.6
10
0.7
1900
After January 1, 2008
1.5
0.3
2
0.07
1900
After January 1, 2012
0.15
0.15
1
0.03
1650
California Air Resources Board
Distributed Generation Certification Rule for New Non-Emergency Engines
After January 1, 2003
0.5
0.5
6
1
fuel req.
After January 1, 2007
0.07
0.07
0.1
0.02
fuel req.
b
b
Airborne Toxics Control Measure for New and In-Use Stationary IC Engines (DRAFT 6/5/03)
New diesel engines > 50 hp
(converted from g/bhp-hr)
c
Baseload power
off-road standards apply
0.03
Emergency power
Exisiting diesel engines > 50 hp
off-road standards apply
0.44
d
c
Baseload power
NOx, CO and VOCs not to increase > 10% to meet PM limits
Emergency power
NOx, CO and VOCs not to increase > 10% to meet PM limits
e
e
0.03
f
d
1.48
Texas - new engines
Before January 1, 2005
less than 300 hrs/yr
21
1.65
more than 300 hrs/yr
3.11
0.47
After January 1, 2005
less than 300 hrs/yr
21
0.47
more than 300 hrs/yr
3.11
0.14
a
The OTC Model Rule offers three compliance options: 1) meet the NOx emission limit specified above, 2) meet a
percentage reduction of NOx emissions specific to the type of engine, and 3) purchase of NOx allowances.
b
PM emission limit corresponding to natural gas with fuel sulfur content of no more than 1 grain/100 standard cubic foot.
c
Engine must meet model year off-road compression-ignition engine standards, or Tier 1 off-road certification standards.
d
Allowable hours of operation increase as the emissions factor of an engine decreases.
e
Many engines may require control technology to meet the PM limits set in this rule, and these technologies must not
increase the emisisons of NOx, CO or VOC by more than 10%.
f
Engines can meet this standard or reduce PM emissions by 85%.
105
B. New England Demand Response Initiative and Efforts to Improve
Data Collection and Coordination with ISOs
Launched in 2001, the New England Demand Response Initiative (NEDRI) has provided
a useful venue for exploring environmental concerns associated with current federal and
regional efforts to promote more robust demand response capabilities in deregulated
wholesale electricity markets. NEDRI itself was designed as a multi-stakeholder process
that aimed to develop specific policy recommendations on a whole host of demand
response issues. While this effort was targeted from the outset to the needs and concerns
of the New England region specifically, the Federal Energy Regulatory Commission
(FERC) has since indicated that it will look to NEDRI for policy guidance in developing
demand response provisions in its forthcoming rulemaking on Standard Market Design
for competitive wholesale markets nationwide. NEDRI’s recommendations are expected
to be released in a final report due for completion during the summer of 2003.
Meanwhile, a number of specific recommendations concerning eligibility for
participation and information collection requirements in the context of the New England
ISO’s summer 2003 demand response programs have already been finalized by
participating stakeholders and submitted to FERC for review.
Recognizing that states will need to update current air permitting requirements to handle
a greatly expanded reliance on distributed generators, and indeed, that many are already
in the process of doing so, the NEDRI recommendations essentially create two new types
of requirements. First, they establish more explicitly that compliance with current air
permitting requirements is a necessary pre-condition for eligibility to participate in ISOsponsored demand response programs.68 Accordingly, the New England ISO will begin
requiring demand response providers (including third-party aggregators) to affirm that, to
the best of their knowledge, applicable permits are on file for any on-site generators
expected to operate as part of the demand response capability being registered. If an
engine does not require an air permit, the owner is obligated to obtain a waiver from the
state permitting authority and to provide basic information about the engine (e.g. size,
age, model, rated fuel input, etc.).
The second type of requirement endorsed by NEDRI relates to information collection and
coordination between the ISO and state air agencies. Specifically, the New England ISO
has committed to providing state air agencies with a summary report on the number of
hours and dates that participating units actually ran and the megawatt hours of electricity
produced during demand response events at the end of each summer season. To allow a
full and detailed assessment of environmental impacts from these programs, future
reports of this nature will ideally include specific information about ownership, type,
size, and location of engines, control technologies in place and estimated emissions, type
of fuel used (including sulfur content), amount of electricity produced, and hours/dates of
demand response events. As discussed at length in the NEDRI process, improved
information is crucial for state air programs to evaluate what kinds of control
68
Note that this basic approach has already been adopted by PJM – the ISO that serves much of the MidAtlantic region (see Footnote 8) – with FERC approval.
106
technologies and strategies (including perhaps an allowance-based market based
approach such as the OTC NOx Budget program) may be needed to ensure that adequate
and cost-effective environmental protection accompanies the increased use of demand
response resources. Enhancing and formalizing information exchange between the ISO
and air regulators can go a long way toward addressing and resolving lingering concerns
about the environmental impacts of distributed generation and toward ensuring that state
permitting and enforcement programs are providing effective protection against
unwanted environmental outcomes.
C. Current State Activities Related to the Environmental Regulation of
Distributed Generators in the Northeast
As noted earlier in this chapter a number of Northeast states have already begun efforts to
update permitting requirements and other regulations pertaining to distributed generators,
including diesel engines. A short summary of each state’s activities (updated as of May
2003) is presented below.
1. Connecticut
As described in Chapter II and elsewhere, Connecticut in 2002 introduced a new General
Permit for Distributed Generation – applicable to all engines over 50 hp (37 kW) – to
help respond to resource adequacy concerns in the southwest Connecticut load pocket.
This permit expires at the end of 2003 and Connecticut is currently developing a separate
regulation to be in place in spring 2004. This new regulation is likely to be a selfimplementing permit-by-rule applicable to distributed generators in the entire state, and
will likely set technology-neutral emissions limitations that become more stringent over
time. The Connecticut air bureau also co-chaired the RAP collaborative that developed
the model emissions rule described earlier in this chapter. The RAP model rule is being
used for guidance in Connecticut’s current development of new regulations for
distributed generation. Meanwhile, the state currently prohibits emergency engines from
participating in the New England ISO price response program.
2. Maine
In Maine, emergency engines are licensed to operate during “sudden and reasonably
unforeseeable events beyond the control of the source” for a maximum of 500 hours per
year. Non-emergency engines over 500 kW in Maine must be licensed and all diesel
engines are required to use on-road fuel. Regulators in Maine do not currently view
distributed generation as a significant problem, but they are experiencing an increase in
the number of companies requesting permits to install engines for primary power. Thus,
peak and baseload engines represent a greater source of potential emissions increases
than emergency engines in Maine. Accordingly, the Maine Department of Environmental
Protection has stated that it will require selective catalytic reduction (SCR) controls for
any engine with potential NOx emissions in excess of 20 tons per year. At this time, no
107
generator has triggered this requirement. Currently, regulators in Maine are beginning a
process to develop a rule that will address distributed generation.
3. Massachusetts
Air quality officials in Massachusetts are developing new regulations to include a larger
number of distributed generators in the state’s permitting system. At the same time, the
state is considering streamlining the pre-construction review process – which is currently
conducted on a case-by-case basis for all engines above particular size thresholds – to
introduce a certification and permit-by-rule compliance option for non-emergency
engines. Certification requirements will also be added to the existing permit-by-rule
program for emergency engines. Together with Connecticut, Massachusetts air officials
co-chaired the RAP model rule collaborative. They are now considering adoption of
regulations based on the RAP model rule, including the extension of permit requirements
to smaller engines.
Massachusetts anticipates that emissions requirements for non-emergency engines will
continue to be more stringent than those for emergency engines and that only engines
meeting the more stringent non-emergency requirements will be eligible to participate in
the New England ISO’s price response programs. In general, Massachusetts considers the
widespread use of emergency engines for stand-alone generation to be quite limited,
given the unfavorable economics of operating these units as a substitute for grid-supplied
electricity. Finally, the Massachusetts Department of Telecommunications and Energy
recently concluded a stakeholder process to recommend new interconnection standards,
policies and procedures to support greater development of distributed generation
resources. Those policies are now being considered for adoption by the state.
4. New Hampshire
As described in the next section (Section D) of this chapter, New Hampshire introduced
an emissions fee system in 2001 to promote reduced NOx emissions from non-emergency
generators. The fees do not apply to very small generators or to emergency generators.
They range in magnitude from $400-$800 per ton of NOx emissions, depending on
whether emissions occur during the summer ozone season (May 1 to September 30), and
are scheduled to increase to $500-$1,000 per ton in the next few years. At this time, New
Hampshire is not working on additional regulations to address distributed generation.
5. New Jersey
New Jersey plans to propose tighter NOx restrictions for non-emergency reciprocating IC
engines, consistent with the OTC model rule. Accordingly, new or modified engines
larger than 50 hp (about 37 kW) will require a permit and will need to meet more
stringent state-of-the-art emissions limits for NOx and particulate emissions. At the same
time, New Jersey plans to raise the permitting size thresholds for clean distributed
108
generators, as proposed in the OTC Model Rule for Distributed Generation, so that only
larger units of this type will need to obtain permits.69
New Jersey’s current permitting restrictions preclude the operation of emergency
generators in anything but outage situations (i.e. “only when the primary source of energy
has been rendered inoperable by circumstances beyond the control of the owner or
operator of the facility”). However, the state is considering proposing changes that would
allow emergency generators to operate to avert imminent outages – for example, when a
voltage reduction alert is called by PJM. This would be similar to the emergency demand
response exceptions that have been introduced in New York and some parts of New
England.
6. New York
Since the summer of 2002, the New York DEC has prohibited the use of all diesel
engines and all emergency engines in the New York ISO’s price responsive program.
However, emergency engines are eligible to participate in the NY ISO’s emergency
demand response program, subject to certain requirements including limits on total
operating hours.
The NY DEC began a stakeholder process as part of its Distributed Generation
Rulemaking Project in 2001. The proposed rule includes RACT-level output-based limits
for NOx and CO emissions from new distributed generation sources. The proposed PM
emissions standards are input-based and have been in place for all oil-fired engines since
1972. The PM emissions standards will apply to existing engines in 2007, and the NOx
and CO standards will apply to existing engines in 2008. Finally, the state is proposing to
require the use of ultra-low sulfur fuel in diesel generators and to phase in new emissions
standards over the next three years. New York air officials hope to have a rule finalized
by early 2004.
7. Rhode Island
Rhode Island has plans to propose a new “general permit” or “permit-by-rule” for
emergency generators in 2004. This may include engines smaller than the current 500 kW
threshold. For other non-emergency engines, the state has considered options such as
including smaller engines or adopting the RAP model rule, but has not yet made a final
decision. Rhode Island air officials are also interested in observing other states’
experience with adopting the OTC model rule before determining a final course of action.
Currently, emergency engines in Rhode Island are allowed to operate only “when the
primary power source is inoperable.” As such they are precluded from participating in
any interruptible power service agreements, including the New England ISO demand
69
The OTC Model Rule for Distributed Generation recommends exempting the following engines from
permitting requirements: All fuel cells fueled by hydrogen, fuel cells smaller than 5 MW fueled by
methane, all remaining fuel cells smaller than 500 kW, and microturbines smaller than 500 kW fueled by
natural gas and certified to emit less than 0.4 lb/MWh of NOx.
109
response programs. To become eligible for participation, any engine over applicable
permitting size thresholds would be required to utilize Best Available Control
Technology (BACT). BACT performance requirements for a given unit would vary
depending on expected hours of operation, but would at a minimum require the use of
ultra-low sulfur fuel for diesel engines.
8. Vermont
Currently, Vermont allows emergency engines to operate only for testing (maximum of
200 hrs/yr) or in true emergency situations (i.e. outages). There is no restriction on hours
of operation in the event of a true emergency. The state has no plans at present to allow
emergency engines to operate to avoid emergencies (e.g. through emergency demand
response programs). Non-emergency engines may, of course, participate subject to NSR
requirements and any other state-imposed fuel or operational restrictions. Meanwhile, the
state is planning to revise its emissions standards for both emergency and non-emergency
diesel generators to bring them in line with federal standards for non-road diesel engines.
Requiring SCR NOx control for some high-use engines is also an option being
considered, particularly for engines used in ski areas.
D. Control Technology, Cost and Other Policy Considerations
As discussed at length in Chapter VI, a number of control technologies exist that can
substantially reduce diesel engine emissions of pollutants such as nitrogen oxides (NOx),
carbon monoxide (CO), particulate matter (PM) and hydrocarbons or volatile organic
compounds (VOCs). The primary control options include diesel particulate filters,
oxidation catalysts and selective catalytic reduction (SCR) for NOx control. Generally,
these technologies are well developed and have been successfully demonstrated. Thus the
chief issues when evaluating emissions control policies for diesel IC generators –
particularly in the case of smaller engines – are likely to be issues of cost and costeffectiveness, rather than technical feasibility. If an engine runs for only a few tens of
hours in a year, a given control technology will likely remove only relatively small
amounts of emissions (say, compared to a central-station power plant) for a given capital
cost.70 Not surprisingly, data from the case studies described in Chapter VII indicate that
cost effectiveness – as measured by the conventional metric of tons removed per dollar of
control cost – improves as operating hours increase.
In this context, innovative regulatory approaches can provide attractive alternatives to
mandating end-of-the-pipe emissions control technologies. For example, in late 2001,
New Hampshire imposed NOx emissions fees (in dollars per ton) on small diesel
generators in the state in an effort to reduce NOx emissions from these sources. The fees
apply only to NOx emissions; in addition, small engines (with NOx emissions below 5
tons per year) and emergency generators are exempted. New Hampshire’s current fees
70
It should be emphasized that the relatively high capital costs currently associated with many available
control systems can be expected to decline substantially with manufacturing economies of scale and as
more engines are retrofitted, creating a more competitive market for vendors.
110
range from $400 per ton of NOx emissions in the non-ozone season up to $800 per ton
during the ozone season. They are scheduled to increase to $500 per ton and $1,000 per
ton, respectively, in 2005. Similar emissions incentive programs could be applied
elsewhere in the Northeast to promote less polluting engines. The resulting revenue
stream could be used for control technology development or demonstration programs,
retrofit efforts, or to support more advanced distributed generation options including
inherently low or zero emission renewable and fuel cell alternatives. Finally, another
option would be to require distributed generators participating in ISO or utility-sponsored
demand response programs to obtain pollution allowances. This approach is most readily
implemented where an allowance trading program already exists for other sources, as in
the case of the existing OTC NOx budget program which caps ozone-season NOx
emissions from power plants and large industrial boilers in the Northeast.
In developing future emissions standards and permitting requirements for diesel
generators, states should consider the new federal standards recently introduced for nonroad diesel engines used in farming, construction, and industrial activities.71 The
standards require substantial reductions in NOx and PM emissions (on the order of 90%
from current levels) to be achieved in new engines by 2014. In addition, they establish
fuel content requirements (i.e. sulfur caps).72 New fuel requirements are necessary
because the control technologies necessary to meet more stringent emissions standards
(e.g. diesel particle filters for PM and NOx adsorbers73 for NOx control) require lowsulfur fuel. Importantly, low-sulfur fuel will provide immediate emissions reductions in
the existing engine fleet, in addition to any reductions that are gradually achieved by the
introduction of new and cleaner engine models. In addition, emulsified diesel fuel74 holds
considerable promise as a cost-effective strategy for reducing both PM and NOx
emissions. Some recent commercial diesel products utilizing an emulsion formulation
with 20% water content have been verified by the California Air Resources Board to
provide average PM and NOx emissions reductions of 63% and 14%, respectively, when
used in on-highway vehicles.75 The transferability of this fuel-based control option to
stationary diesel engines needs to be investigated. Meanwhile, fuel standards (especially
related to sulfur content) could be adopted for stationary diesel generators to achieve
near-term emissions reductions – and indeed a number of northeastern states have already
taken this step.
As in the case of stationary engines used to generate electricity, emissions standards for
new non-road engines will take considerable time to penetrate, given the slow turnover of
71
EPA promulgated its rule Control of Emissions of Air Pollution from Nonroad Diesel Engines and Fuel
on April 15, 2003 (40 CFR Parts 69, 80, 89, 1039, 1065, and 1068). The rule was published in the Federal
Register on May 23, 2003 (http://www.epa.gov/fedrgstr/EPA-AIR/2003/May/Day-23/a9737a.htm).
72
Specifically, EPA is requiring that the allowable sulfur levels in nonroad diesel fuel be reduced by more
than 99% (from 3400 ppm to about 15 ppm by the year 2010).
73
NOx adsorber technology is applicable to new engines, but not as a retrofit option for existing engines.
74
Generically, the term emulsion refers to the suspension of small globules of one liquid in a second liquid
with which the first will not mix. In this case, the emulsion would consist of diesel globules suspended in
water.
75
Information on emulsified diesel is taken from NESCAUM’s Draft White Paper Status Report on Clean
Mobile Source Diesel Initiatives in the Northeast States’ and Canadian Provinces, April 2003.
111
the existing fleet. As a result, efforts are also underway to develop and apply retrofit
emission control technologies for on-road and non-road engines, many of which are
likely to be similarly applicable to existing stationary diesel generators.76 For example, a
number of voluntary programs are being used in the Northeast to test the operation of
diesel particulate filters and oxidation catalysts in on-road and non-road mobile source
applications (e.g. trucks, buses, construction equipment, etc.). There is no technical
reason why these options cannot be successfully applied and should not be tested on
existing stationary diesel engines.
E. Policy Recommendations for the Northeast States
Based on the findings of this report and other related initiatives, NESCAUM has
developed a number of policy recommendations for Northeast states interested in
regulating diesel generators specifically and promoting cleaner distributed generation
alternatives more broadly. In general, our policy recommendations fall into three
categories:
•
•
•
Updating emissions standards and air permitting requirements
Regulating use of diesel generators in demand response programs
Improving regional coordination and data collection
Each of these categories of recommendations is discussed in more detail below.
1. Updating Emissions Standards and Permitting Requirements
State air regulators increasingly recognize that current permitting requirements may need
to be updated to manage possible adverse impacts associated with the use of diesel IC
engines, whether as part of formal demand response programs or in response to changing
market conditions. The need for new regulation may be particularly acute for smaller
units that fall below current permitting thresholds. Current exemptions for emergency
engines may continue to be appropriate, provided these engines continue to be barred
from participation in economic demand response (or “price response”) programs. To the
extent that emergency generators are allowed to participate in emergency demand
response programs, current operational limits or other restrictions may need to be
revisited in light of the possibility that such programs may be triggered much more
frequently than actual power outages (particularly in existing load pockets and, more
broadly, as demand catches up with supply in the Northeast over time). At a minimum,
states should consider requiring that emergency generators eligible to participate in
emergency demand response programs be operated on “clean diesel” or low sulfur fuel,
which can reduce PM emissions (and possibly NOx emissions) by as much as 10-20% at
a very reasonable cost. In addition, such fuels can provide substantial sulfur dioxide
(SO2) emissions reductions.
76
Note that while states are preempted from regulating mobile onroad and nonroad diesel engines, they are
not preempted from setting standards for new and existing stationary engines. Hence, states can move to
adopt emissions standards for distributed generators even absent federal action on this issue.
112
States should also consider whether retrofit requirements can be cost-effectively applied
to existing peak-shaving or baseload diesel generators that stand to increase their output
under formal price response programs or as a result of real-time pricing and other market
signals. Additional retrofit or other requirements (such as those proposed in the OTC
model rule) may also be appropriate for otherwise uncontrolled emergency generators
that might begin operating more frequently under ISO-sponsored emergency demand
response programs. As noted previously, efforts aimed at reducing emissions impacts
from the existing generator base can benefit from parallel efforts underway for mobile
diesel engines (both on-road and non-road) and from the availability of low-sulfur fuel. In
evaluating the need for new emissions control requirements for existing diesel generators,
states should consider that the control of particulate emissions may be both more
important from a public health perspective and more cost-effective from a control
technology perspective. For reasons discussed elsewhere in this report, NOx emissions
from small diesel generators are likely to remain relatively modest in the context of
states’ overall ozone attainment planning obligations in most areas. However, even
infrequent short-term increases in particulate emissions from these types of generators
can impose substantial health risks on local communities (and especially on susceptible
individuals) in the vicinity. In addition, it appears that particulate filters and oxidation
catalysts may be relatively inexpensive and easy to retrofit on existing engines compared
to NOx controls such as SCR.
To ensure that the region can take advantage of the multiple system benefits of
distributed resources in the future, states should consider adopting regulations for new
distributed generators. By developing standards that are fuel-independent and outputbased (e.g. lb/MWh), states can promote new technologies and improved efficiencies in
future distributed generation technologies. In particular, designing regulations to account
for useful thermal as well as electrical output can be used to support increased penetration
of highly efficient combined heat and power systems in commercial and industrial
applications. In developing requirements for new generators, states should consider
including, at a minimum, rigorous performance standards for emissions of particulate
matter (preferably fine PM or PM2.5), NOx, toxics and carbon monoxide (CO). In
addition, certification procedures and in-use testing requirements, as well as other
provisions, may be needed to promote proper maintenance and ensure effective
implementation and enforcement. For example, the new standards adopted by Texas in
2001 require re-certification of engine emissions after 16,000 hours (but not more than
once every 3 years).
Finally, states may wish to consider providing incentives for truly advanced, ultra-low or
zero-emissions technologies in the permitting process. For example, Texas waives
permitting fees for distributed generator units that have certified NOx emissions less than
10% of the required standards, as long as total generating capacity is less than 1 MW.
Similar incentives could be used to promote highly efficient, low-emitting fuel cell
technologies and zero-emissions renewable options (e.g. wind or solar). Here again, such
programs should be designed to account for the increased efficiency of combined heat
and power systems, which capture the heat generated by the electricity conversion
process to provide useful thermal energy (typically in the form of steam).
113
2. Regulating Use of Diesel Generators in Demand Response Programs
States should evaluate the need for updated regulations to ensure that the use of diesel
generators in demand response programs sponsored by ISOs or by local distribution
utilities does not create unintended public health and air quality impacts. In general, it is
probably appropriate to draw a clear distinction between economic (price-driven) demand
response programs and emergency demand response programs. Different – and more
stringent – environmental eligibility criteria may be appropriately applied to economic
(price-driven) programs because their potential environmental impacts are substantially
greater than those of emergency programs, which, by definition, are invoked only rarely
and then usually for short periods of time (i.e. on the order of tens and not hundreds of
hours per year). An additional important justification for more stringent environmental
controls in price response programs is the fact that participants may receive substantial
compensation for any on-site generation they provide on high-demand days when
electricity may be valued at prices ten times or more the average price. In light of these
payments, additional pollution control investments may be justified, even if their costeffectiveness – as measured in dollars per ton of pollutant removed – is less favorable
than the typical cost-effectiveness for emissions controls applied to central station power
plants.
As was noted in the above recommendations concerning permitting requirements and
emissions standards, modest retrofit requirements and/or fuel quality standards may also
be appropriate for emergency generators participating in emergency demand response
programs, particularly if they are likely to accumulate hours of operation at or near
current permit limits. In addition, it is important that the definition of what constitutes an
emergency be clearly defined (preferably on a consistent basis region-wide, as discussed
in the next section). To date, most northeastern state air regulators have defined
“emergency” to correspond to a particular step in ISO operations, typically necessitating
an actual (as opposed to anticipated) call for voltage reductions.
For economic or price-driven demand response programs, states should consider
restricting eligibility to distributed generators that meet minimum emissions performance
standards. In addition, further restrictions or requirements specific to diesel engines may
be appropriate in some cases, especially in ozone and PM non-attainment areas or in
areas with particular local air quality concerns (e.g. environmental justice issues or a high
concentration of other diesel exhaust sources). In any case, environmental eligibility
requirements for all types of ISO or utility sponsored demand response programs must be
clearly communicated to intermediary parties and potential program participants. Ideally,
future updates of state emission control requirements and permitting programs should
ensure that it is easy to determine and document a given participant’s eligibility for
different types of demand response programs. For their part, the ISO and distribution
utilities should continue to work with state regulators as permitting requirements evolve
to ensure that generation owners have accurate, up-to-date information about permitting
requirements and reporting obligations.
114
Meanwhile, the recommendations of the NEDRI process concerning the obligation of
demand response providers to check their permit status, to obtain a waiver if they do not
require a permit and to provide basic information on any on-site generators (see Section
B of this chapter), together with the responsibility of the ISO to provide specific afterthe-fact information on program outcomes, provide a useful model for other regions and
can serve as an important environmental backstop and source of information as state
policies evolve.
3. Improving Regional Coordination and Data Collection
A regional approach to implementing many of the policy recommendations described in
this chapter can improve effectiveness, reduce confusion and level the competitive
playing field for regulated entities, and minimize regulatory and administrative burdens
for states facing formidable resource constraints at this time. At present, state permit
requirements vary widely and are perceived as complex and confusing by many
economic regulators, system operators and generation owners. More consistency would
make it easier for ISOs, the regulated community and other stakeholders to work together
in ensuring compliance. Additionally, the adoption of regionally consistent emissions
standards for new generators would send a much more powerful signal to manufacturers
of distributed generation technologies and would increase the probability that compliant
models are brought to market in a timely manner.
In the Northeast, NEDRI has recommended some important initial steps toward greater
regional consistency. In the interests of continued progress in this direction, NESCAUM
recommends the Northeast states take steps to promote further regional consistency in:
(1) regulating the emissions performance of new and existing generators and (2)
establishing the eligibility of different generators for various types of demand response
programs. In the meantime, efforts already underway to clarify current state requirements
and to facilitate permit checking and information collection by state air agencies and the
ISO should continue and be strengthened.
In addition to promoting regionally consistent permit rules, demand response eligibility
and data collection requirements, the Northeast states should build on this initial effort to
develop more reliable and comprehensive inventories of internal combustion engines in
the region. Greater regional consistency in state record-keeping practices would be
helpful in this regard and could improve states’ ability to use available permit
information to assess environmental impacts and identify policy priorities. Eventually,
the development of a region-wide, user-friendly database of distributed generation
resources may be feasible. The ability to query a single database for specific information
on large numbers of individual generators would be extremely helpful to state air
regulators and to other policy makers interested in promoting environmentally
responsible distributed resources more broadly.77
77
For example, a database could be designed to include key pieces of information, including: (1) permit
number, (2) owner name, address and contact information, (3) engine location, (4) make and model of
engine, (5) engine size, (6) engine purpose, (7) estimated hours of operation each year, (8) eligibility for
demand response programs, (9) fuel type and (10) emissions rates and control technology.
115
F. Conclusions
This report represents a first and necessarily incomplete effort to develop more
comprehensive estimates of the population and associated emissions potential of existing
diesel generators in the Northeast. To refine these results in the future, regional data
collection mechanisms must be improved. In the meantime, states should consider
adopting new requirements and policies aimed at managing any adverse environmental
impacts associated with increased reliance on the existing base of distributed generators
and promoting a transition to cleaner distributed generation options in the future. The
various policy recommendations described in this chapter are re-summarized in Table
VIII-2. Most of these recommendations are directly aimed at new and existing generators.
However, it is worth emphasizing that a number of other state policies and regulations are
likely to directly or indirectly affect the future use and composition of distributed
generation options in the Northeast. Examples include utility pricing reforms,
interconnection standards and net metering requirements, transmission planning and
resource adequacy determinations, information disclosure and emissions portfolio
requirements, emissions cap-and-trade programs and other policies. The interaction of
these policies and programs with policies more directly related to distributed generation
is therefore an important and ongoing consideration for states as they seek to promote a
more reliable, less polluting, less costly and more efficient system for meeting the
region’s electricity needs.
116
Table VIII-2
Summary of NESCAUM Recommendations
Updating Emissions
Standards & Permitting
Requirements
•
•
•
•
Regulating Use of Diesel
Generators in Demand
Response Programs
•
•
•
•
•
Improving Regional
Coordination and Data
Collection
•
•
•
Review adequacy of current permitting size thresholds and
requirements, especially in light of new health concerns associated
with localized exposure to diesel exhaust.
Update requirements for existing peak, baseload and emergency
generators accordingly, especially for those units eligible to participate
in ISO or utility-sponsored demand response programs.
Consider additional fuel requirements (e.g. use of low or ultra-low
sulfur fuel) for diesel generators, especially for operation in demand
response or other non-emergency applications.
Adopt stringent output-based emissions standards for new distributed
generators, which – in the case of combined heat and power systems –
appropriately account for useful thermal, as well as electrical output.
Limit participation in non-emergency economic (price-driven) demand
response programs to generators that meet minimum emissions control
requirements (e.g. BACT-level controls, RAP model rule, etc.)
Limit participation of emergency generators to emergency demand
response programs, subject to additional requirements as deemed
appropriate under recommendations above.
Clarify regionally consistent definition of “emergency”.
Consider appropriateness of restrictions and/or additional emissions
control or operating requirements (e.g. fuel sulfur requirements) for
diesel generators participating in demand response programs.
Continue implementing NEDRI recommendations concerning the
obligation of demand response participants to verify permit status and
provide unit specific information, together with the ISO’s obligation to
provide detailed information about program outcomes on a regular
basis. The information provided must be specific enough to allow air
regulators to make a reliable assessment of associated environmental
and public health impacts.
Promote more regional consistency in permitting requirements and
emissions standards for new and existing distributed generators.
Promote more regional consistency in state record-keeping practices,
with the aim of eventually developing integrated, user-friendly
information databases.
Continue to develop and refine inventories of generator population and
potential emissions impacts. Promote regional consistency in related
policies (e.g. interconnection standards, pricing policies, other air
regulatory programs, etc.)
117
Appendix A
Power Systems Research Methodology
PARTSLINK™ METHODOLOGY
BACKGROUND
Over the past 23 years Power Systems Research has maintained a database known as
PartsLinkTM. PartsLinkTM utilizes a mathematical model to estimate the number of engine
powered products in service in the United States. This model is developed through the use of
factual data and survey results developed and compiled by Power Systems Research on a
continuing basis. Application populations are distributed geographically based upon selected
economic factors from the Bureau of Census, County Business Patterns Survey.
Key elements of the data include:
• A continuing record of shipments from U.S. factories and imports from foreign suppliers
• Exports of U.S. product equipment
• An attrition model utilized to estimate retirement of engine powered products based upon:
¾ Estimated engine life
¾ Annual hours of utilization
¾ Intensity of utilization – load factor
These factors are utilized to calculate retirement rates and estimate the resulting number of
products remaining in operation.
Estimated Engine Life
Because accurate data on actual engine life is not available, Power Systems Research has
developed an estimating methodology, which incorporates a wide variety of identifiable engine
characteristics to predict the useful life of an engine at maximum continuous output. These
factors include:
•
•
•
•
•
•
•
•
•
Engine horsepower
Rated speed
Number of cylinders
Displacement
Aspiration
Engine weight
Configuration
Bore
Stroke
These variables combined with constants, developed by Power Systems Research when
comparing projected engine life to a few known benchmarks, have been utilized to calculate a
projected engine life for every engine in the database. Engine life is expressed in the average
number of hours an engine will operate at the maximum continuous rated output for the engine.
The product of this horsepower and the number of hours describe life in horsepower hours.
Because normal operation does not involve operation at full output, our assumption says that
engine lifetime in actual hours is extended when the engine is operated at less than full output,
A-1
but the number of horsepower hours will always be constant. We have assumed that this lifetime
will be a statistical mean and that a normal distribution can be used to describe all retirements.
Activity Levels
The hours per year experienced by an owner will vary considerably, but generally are similar for
any given product application. In our survey, we ask for annual operating hours. In our model we
use the mean hours per year for each application, which results from response to this question.
Fuel Consumption
The average output or load factor is typically similar within an application. Because users are
usually not able to measure load factor (the average percentage of maximum output at which
they are operating), a good indicator is the amount of fuel they consume. Fuel will vary almost
directly with the amount of horsepower produced. We ask for annual fuel consumption and then
compare that response to fuel which would be consumed at maximum horsepower. The result is
load factor. In many cases, respondents are not able to easily estimate annual hours of operation
or fuel consumed annually. In those cases we have asked respondents to estimate hours per
week and fuel consumed per week. This also required questions to determine if use was
seasonal, length of season, and use in off-season. The results were then projected over 52
weeks to get an annual result.
Survey Research
In order to develop reliable parameters for aftermarket indicators, Power Systems Research has
developed an on-going survey of owners of engine powered products. The objectives of this
survey include development of:
•
•
•
•
Mean product lifetime
Annual hours of operation
Typical load factor
Replacement frequency for key components
This survey is conducted among randomly selected owners of each type of engine powered
equipment in each market region, and asks users a rather simple set of questions:
•
•
•
•
•
•
•
•
•
Type of equipment operated
Equipment manufacturer
Equipment model
Age of the equipment
Cumulative hours
Hours operated during the past 12 months
Fuel consumed during the past 12 months
Engine installed, if available
Year of manufacture, if known
Typically we have found it necessary to establish these benchmarks through completion of 100
interviews with owners of each type of equipment in each market region. NOTE: “Equipment or
Product Type” is denoted by application [see OeLink Product Guide], fuel, and power range.
Over the years we have implemented a number of “rules” to facilitate data processing and
storage rather than statistical reliability. Whenever possible, we benchmark our data against
widely accepted authoritative sources. Our effort is to establish credible, statistical reliability for
the operating characteristics developed during the course of our survey.
A-2
Geographic Distribution
The geographic distribution of generator sets in service is accomplished by matching ownership
and application norms to economic data provided by the Census Bureau, County Business
Patterns database. The ownership and application norms have been developed over the past
20+ years by developing a profile for owners of each type of equipment. These profiles consist
of a correlation between any combination of up to 22 economic, geographic, demographic and
meteorological factors. For each county or combination of counties selected the profile for owners
of equipment in any application and power range is compiled based on Census Data [Standard
Industrial Classification (SIC) and employee size] data and the strength of this profile is compared
to the national profile for that same application. The proportionate allocation of equipment in that
application and power range is then assigned to that county. The association can be made for
application, power rating, engine or equipment brand and model, and age distribution. With each
level of specificity the statistical reliability deteriorates somewhat. The profile norms are simply
tell us the probability that a certain type of business entity or consumer will own a specific type of
equipment. We then complete a normal distribution over the profile interval to determine what
portion of the national population is located in any specific area.
ADDITONAL NOTES ON THE OWNER SURVEY
Each year, our survey is directed to a random sample of businesses and consumers. In each
case we identify engine powered products owned by the respondent. We then collect operating
data for that respondent along with demographic and economic data. This information then
becomes the basis for projecting the geographic allocation of engine powered products. We look
at the distribution of generator sets among businesses by SIC and by employee size as well as
location. From this information we are able to establish a correlation from which we can project
the population across the entire nation. For example we may find that metal fabricating
companies [SIC 331] with between 400 and 600 employees own 14% of the generator sets
between 200 and 300 kW. We know from our sales record and attrition that there are 150,000
units nationwide in this power range and thus 21000 units owned by companies of this size and
SIC. We can find from Census data that there are 63000 such companies nationwide so we can
project that there will be 1 generator in this power range for every three companies of this type. If
there are 60 companies of this description in an area, we would then project that there are 20
generators of this size owned among them.
This methodology and completion of more than 200,000 interviews over the past 20 years has
allowed us to construct a matrix for SIC vs. Product type. The data contained in this matrix is the
nationwide incidence of ownership for each product type by companies within each SIC and
employee size or by consumers. Further derivatives of this matrix such as smaller geographic
areas [down to the county level] more specific SIC’s and/or more specific product specifications
can be compiled but of course the statistical reliability declines as the information becomes more
granular.
APPLICATION TO THE NESCAUM SURVEY
For the purpose of targeting survey sample, such as in the NESCAUM survey, by type of engine
used and product type, we look at the nationwide incidence for each SIC and company size. We
compare that to a tabulation of companies by SIC and size in the target area. This comparison
gives us a first estimate of how many units we will find in the area and how many owners we will
need to contact in order to find those units. In most cases some of the owners we have previously
contacted through our normal surveys are found in this area and we supplement the verified
owner list by directing our survey first to the highest incidence group and successively to lower
probability owners.
The target survey sample is first constructed to draw names from enough combinations of SIC’s
and employee sizes to reach our quota [in the NESCAUM case identification of 90% of projected
A-3
population of stationary generator sets]. In this case the estimation was somewhat more difficult
because our product types do not delineate between stationary and portable. When we have
exhausted our sample and not reached our quota we continue to draw sample from those
categories which have demonstrated the best success rate until we have either totally exhausted
all sample or reached our quota.
In this case we estimated that in general we should target all SIC’s in the target areas for which
sample was available and having more than 50 employees. In several cases we exhausted the
sample before reaching our quota and thus we continued sampling to companies with less than
50 employees. Our review indicates that there were some deficiencies in the list supplied by
American Business Lists. Several SIC’s which subsequently were found on the list of installed
and permitted generators in the target area were missing from the American Business List
information. As a result our compilation is probably over representative in some SIC’s and under
representative in others. Nonetheless, the results and incidence of ownership in the target areas
were similar enough to the national pattern to make statistically valid estimates of the installed
generator set population.
One exception is derived from an anomaly in our attrition model. When an engine is projected to
have reached two times the mean lifetime expectation for that engine it is dropped from our
tabulation – we assume it has been retired. This is often not the case – especially in the case of
large stationary generator sets. These gene sets typically are dropped after a lifetime of 30 to 40
years but in fact many survive for much longer because they do not operate for any appreciable
amount of time per year.
Among lessons learned from this examination of Fairfield County and the New York City
metropolitan area, it appears that the most important is the incompleteness in some sectors of the
ABL databank – specifically utilities and government agencies. This is not particularly surprising
since the primary source of ABL information is Yellow Pages directories. We have found similar
shortcomings with the Dun & Bradstreet listings over the years. A further lesson for us is the
importance of creating a correlation to delineate portable and stationary generators.
In the final analysis, the survey methodology was successful in yielding a credible estimation of
the size and distribution for active stationary generator sets in the area as well as the pattern of
use. It also revealed that with each level of specificity the reliability deteriorates and that, at least
in these cases, list of permitted owners will substantially understate the population.
A-4
Appendix B
Power Systems Research Population Estimates for Natural Gas
Engines
NESCAUM Region
Number Totals
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Emergency Peak Baseload Total
268
268
384
364
2
18
152
152
177
177
268
182
77
9
58
22
30
6
26
248
83
357
1,191
357
116
1,664
Capacity Totals (MW) Emergency Peak Baseload Total
25-50 kW
7
7
50-100 kW
26
25
0
1
100-250 kW
27
27
250-500 kW
59
59
500-750 kW
173
116
50
6
750-1000 kW
1000-1500 kW
72
27
37
7
1500+ kW
47
500
114
661
Total
309
588
129
1,026
Emergency Peak Baseload Total
29
29
34
34
6
6
8
8
23
15
8
3
3
1
11
14
26
93
22
14
129
Capacity Totals (MW) Emergency Peak Baseload Total
25-50 kW
1
1
50-100 kW
2
2
100-250 kW
1
1
250-500 kW
3
3
500-750 kW
16
10
5
750-1000 kW
1000-1500 kW
4
4
1500+ kW
2
23
28
52
Total
19
32
28
78
Emergency Peak Baseload Total
1
1
1
1
2
1
3
2
2
1
5
Capacity Totals (MW) Emergency Peak Baseload Total
25-50 kW
0
0
50-100 kW
0
0
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
4
2
6
Total
0
4
2
6
Connecticut
Number Totals
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Maine
Number Totals
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Massachusetts
Number Totals
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Emergency Peak Baseload Total
41
41
60
59
1
24
24
27
27
32
23
9
0
10
5
5
3
42
13
58
182
56
14
252
Capacity Totals (MW) Emergency Peak Baseload Total
25-50 kW
1
1
50-100 kW
4
4
0
100-250 kW
4
4
250-500 kW
9
9
500-750 kW
21
15
6
750-1000 kW
0
1000-1500 kW
12
6
6
1500+ kW
5
83
25
113
Total
45
95
25
165
B-1
New Hampshire
Number Totals
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Emergency Peak Baseload Total
2
2
5
5
18
2
11
5
9
11
5
25
Capacity Totals (MW) Emergency Peak Baseload Total
25-50 kW
0
0
50-100 kW
0
0
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
35
4
22
10
Total
4
22
10
35
Emergency Peak Baseload Total
70
70
99
94
1
4
37
37
48
48
70
49
21
0
13
6
7
11
97
26
134
315
126
30
471
Capacity Totals (MW) Emergency Peak Baseload Total
25-50 kW
2
2
50-100 kW
7
6
0
0
100-250 kW
7
7
250-500 kW
16
16
500-750 kW
45
31
14
750-1000 kW
0
1000-1500 kW
16
7
9
1500+ kW
21
198
53
272
Total
89
221
53
362
Emergency Peak Baseload Total
123
123
183
169
1
13
85
85
94
94
143
95
39
9
32
11
15
6
5
81
21
107
767
582
136
49
Capacity Totals (MW) Emergency Peak Baseload Total
25-50 kW
3
3
50-100 kW
12
11
0
1
100-250 kW
15
15
250-500 kW
32
32
500-750 kW
91
60
25
6
750-1000 kW
1000-1500 kW
40
14
19
7
1500+ kW
9
163
44
216
Total
409
144
206
59
New Jersey
Number Totals
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
New York
Number Totals
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Rhode Island
Number Totals
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
Emergency Peak Baseload Total
1
1
1
1
4
4
3
11
6
4
3
13
Capacity Totals (MW) Emergency Peak Baseload Total
25-50 kW
0
0
50-100 kW
0
0
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
7
8
5
20
Total
7
8
5
20
Emergency Peak Baseload Total
1
1
1
1
2
2
Capacity Totals (MW) Emergency Peak Baseload Total
25-50 kW
0
0
50-100 kW
0
0
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
0
<1
-
Vermont
Number Totals
25-50 kW
50-100 kW
100-250 kW
250-500 kW
500-750 kW
750-1000 kW
1000-1500 kW
1500+ kW
Total
B-2
Appendix C
Power Systems Research Generator Set Estimator
CALCULATING REGIONAL POPULATIONS
PSR’s generator set population estimates are based on an ongoing nationwide survey in
which a sampling of each North American Industry Classification System (NAICS)
group is made annually. Survey respondents are asked if they own a generator set, its
rating, annual hours of service, installation data and other ownership and operating
questions. On the basis of their responses over the years we have compiled an incidence
factor for each NAICS code. The Census Bureau, Dun & Bradstreet and several other
sources can provide the number of establishments within any geographic region (see for
example http://www.census.gov/epcd/cbp/view/cbpview.html).
Once data on establishments are entered into column C, the total number of generators
estimated to be in service can then be calculated, as shown in the example calculation at
the top of the chart. These estimates should be used only for very broad purposes. In
addition, it should be noted that because of the similarity between stationary and portable
generators the estimates also include portable generator sets (which probably account for
60% or more of the total estimated population and are generally smaller than 300 kW).
Each region is different and the incidence of generator sets, which is based upon national
results, may not be applicable to a local region due to the average size of establishments
within a category, the reliability and cost of the utility electric supply, environmental and
other restrictions, as well as generator engine sales and support infrastructure locally.
This chart originated in an Excel spreadsheet. To set up a similar spreadsheet, copy the
values in columns A, B and D. Data for column C should be imported from the Census
Bureau website. A simple calculation then provides the values for column E (where E =
(C x D)/1,000).
NATIONWIDE - U.S.
A
2 Digit
NAICS
B
C
INDUSTRIAL CATEGORY
EXAMPLE CALCULATION
01
02
07
08
09
10
12
13
14
15
D
E
Gen Set
Incidence Gen Gen Sets
Number of
In Service
Sets / 000
Establishments
Establishments
X
AGRICULTURAL PRODUCTION-CROPS
AGRICULTURAL PRODUCTION-LIVESTOCK
AGRICULTURAL SERVICES
FORESTRY
FISHING HUNTING & TRAPPING
METAL MINING
COAL MINING
OIL & GAS EXTRACTION
MINING & QUARRYING-NONMETALLIC MINERALS
BUILDING CONSTRUCTION-GEN CONTRACTORS
C-1
31,797
14,673
199,281
4,671
2,407
180
1,114
24,425
9,002
308,604
Y
X*Y/1,000
6.4
23.2
8.8
20.8
4.4
25.2
93.6
227.2
45.2
87.2
204
340
1,754
97
11
5
104
5,549
407
26,910
16
17
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
67
70
72
73
75
76
78
HEAVY CONSTRUCTION EXCEPT BUILDING
CONSTRUCTION-SPECIAL TRADE CONTRACTORS
FOOD & KINDRED PRODUCTS MFRS
TOBACCO PRODUCTS MFRS
TEXTILE MILL PRODUCTS MFRS
APPAREL & OTHER FINISHED PRODUCTS-MFRS
LUMBER & WOOD PRODS EXCEPT FURNTR MFRS
FURNITURE & FIXTURES MFRS
PAPER & ALLIED PRODUCTS MFRS
PRINTING PUBLISHING & ALLIED INDUSTRIES
CHEMICALS & ALLIED PRODUCTS MFRS
PETROLEUM REFINING & RELATED INDS MFRS
RUBBER & MISCELLANEOUS PLASTICS MFRS
LEATHER & LEATHER PRODUCTS MFRS
STONE CLAY GLASS & CONCRETE PRODS MFRS
PRIMARY METAL INDUSTRIES MFRS
FABRICATED METAL PRODUCTS MFRS
INDUSTRIAL & COMMERCIAL MACHINERY MFRS
ELECTRONIC & OTHER ELECTRICAL EQUIP MFR
TRANSPORTATION EQUIPMENT MFRS
MEASURING & ANALYZING INSTRUMENTS-MFRS
MISCELLANEOUS MANUFACTURING INDS MFRS
RAILROAD TRANSPORTATION
LOCAL/SUBURBAN TRANSIT & HWY PASSENGER
MOTOR FREIGHT TRANSPORTATION/WAREHOUSE
UNITED STATES POSTAL SERVICE
WATER TRANSPORTATION
TRANSPORTATION BY AIR
PIPELINES EXCEPT NATURAL GAS
TRANSPORTATION SERVICES
COMMUNICATIONS
ELECTRIC GAS & SANITARY SERVICES
WHOLESALE TRADE-DURABLE GOODS
WHOLESALE TRADE-NONDURABLE GOODS
BUILDING MATERIALS & HARDWARE
GENERAL MERCHANDISE STORES
FOOD STORES
AUTOMOTIVE DEALERS & SERVICE STATIONS
APPAREL & ACCESSORY STORES
HOME FURNITURE & FURNISHINGS STORES
EATING & DRINKING PLACES
MISCELLANEOUS RETAIL
DEPOSITORY INSTITUTIONS
NONDEPOSITORY CREDIT INSTITUTIONS
SECURITY & COMMODITY BROKERS
INSURANCE CARRIERS
INSURANCE AGENTS BROKERS & SERVICE
REAL ESTATE
HOLDING & OTHER INVESTMENT OFFICES
HOTELS ROOMING HOUSES & CAMPS
PERSONAL SERVICES
BUSINESS SERVICES
AUTO REPAIR SERVICES & PARKING
MISCELLANEOUS REPAIR SERVICES
MOTION PICTURES
C-2
62,335
484,515
37,214
889
9,927
28,718
30,117
13,586
14,236
131,739
24,519
3,746
20,446
2,600
20,464
14,996
56,908
93,323
31,822
17,397
23,383
60,130
3,215
42,433
144,695
31,737
22,076
15,652
3,234
69,382
81,301
34,700
454,646
173,514
135,342
57,382
298,820
285,774
178,884
266,800
573,049
650,415
129,374
84,012
95,795
27,565
231,299
437,636
8,434
96,145
598,420
546,914
300,181
108,025
41,281
15.2
663.2
34.8
48.8
85.2
14.0
74.0
184.8
338.8
37.2
267.2
394.8
67.2
46.0
65.2
35.6
60.8
49.2
34.0
92.8
21.2
26.8
10.0
15.2
8.8
12.4
20.8
1.4
225.6
105.2
782.4
1,804.8
36.8
45.2
48.8
72.8
101.6
75.6
88.4
24.4
45.2
34.8
78.4
485.2
249.6
115.6
146.0
24.4
22.0
88.4
44.4
67.2
12.4
4.8
24.8
947
321,330
1,295
43
846
402
2,229
2,511
4,823
4,901
6,551
1,479
1,374
120
1,334
534
3,460
4,591
1,082
1,614
496
1,611
32
645
1,273
394
459
23
730
7,299
63,610
62,627
16,731
7,843
6,605
4,177
30,360
21,605
15,813
6,510
25,902
22,634
10,143
40,763
23,910
3,187
33,770
10,678
186
8,499
26,570
36,753
3,722
519
1,024
79
80
81
82
83
84
86
87
89
91
92
93
94
95
96
97
99
AMUSEMENT & RECREATION SERVICES
HEALTH SERVICES
LEGAL SERVICES
EDUCATIONAL SERVICES
SOCIAL SERVICES
MUSEUMS ART GALLERIES & GARDENS
MEMBERSHIP ORGANIZATIONS
ENGINEERING & ACCOUNTING & MGMT SVCS
MISCELLANEOUS SERVICES NEC
EXECUTIVE LEGISLATIVE & GENERAL GOVT
JUSTICE PUBLIC ORDER & SAFETY
PUBLIC FINANCE & TAXATION POLICY
ADMINISTRATION-HUMAN RESOURCE PROGRAMS
ADMIN-ENVIRONMENTAL QUALITY PROGRAMS
ADMINISTRATION OF ECONOMIC PROGRAMS
NATIONAL SECURITY & INTERNATL AFFAIRS
NONCLASSIFIED ESTABLISHMENTS
TOTAL
181,397
1,203,660
505,140
235,724
317,856
8,831
526,080
386,323
21,719
176,429
82,878
15,953
23,520
18,173
24,162
15,073
586,019
12,336,233
C-3
125.2
456.8
2.5
342.8
85.2
21.6
8.4
4.8
5.2
85.2
194.8
104.8
73.2
104.8
21.6
673.6
0.1
22,711
549,832
1,253
80,806
27,081
191
4,419
1,854
113
15,032
16,145
1,672
1,722
1,905
522
10,153
75
1,629,436
Appendix D
Additional Background on Stationary IC Engines and
Emissions
Note: the text of this Appendix was developed by ESI as part of its report to NESCAUM
on emission control technologies. It is included here in the interests of providing
additional background information on stationary IC engines and their emissions.
Internal combustion (IC) engines are used in a variety of stationary applications ranging
from power generation to inert gas production. Both spark ignition and compression
ignition engines are in wide use. Depending on the application, stationary IC engines
range in size from relatively small (~50 hp) for agricultural irrigation purposes to
thousands of horsepower for power generation or natural gas transmission. Often when
used for power generation, several large engines are used in parallel to meet the load
requirements. A variety of fuels can be used for IC engines including diesel, natural gas,
and gasoline among others. The actual fuel used depends on the owner or operator
preference but can be application dependent as well. IC engines can also be run rich, lean,
or stoichiometrically as shown below.
Typical Engine Types and Fuels for Stationary Applications
Rich Burn
Natural Gas
Propane
Gasoline
Stoichiometric
Natural Gas
Propane
Gasoline
Lean Burn
Diesel
Natural Gas
Dual Fuel
The difference between rich, lean, and stoichiometric engine operation lies in the air to
fuel ratio. Stoichiometric engine operation is defined as having the chemically correct
amount of air in the combustion chamber during combustion. Hence, perfect combustion
would result in the production of carbon dioxide (CO2) and water. However, the fact that
perfect combustion is not possible means that even an engine running stoichiometrically
produces hydrocarbon (HC or VOC), carbon monoxide (CO), nitrogen oxides (NOx), and
particulate matter (PM) emissions. A rich-burn engine is characterized by excess fuel in
the combustion chamber during combustion. A lean-burn engine, on the other hand, is
characterized by excess air in the combustion chamber during combustion, which results
in an oxygen-rich exhaust. Diesel engines inherently operate lean, whereas IC engines
that use natural gas, gasoline or propane can be operated in all three modes.
D-1
The three primary fuels used for stationary reciprocating IC engines are gasoline, diesel
(No. 2) oil, and natural gas. Gasoline is used primarily for mobile and portable engines.
Construction sites, farms and households typically use converted mobile engines for
stationary applications because their cost is often less than an engine designed
specifically for stationary applications. In addition, mobile engine parts and service are
readily available, and gasoline is easily transported to the site. Thus, gasoline is an
essential fuel for small and medium size stationary engines. Diesel fuel is also easily
transported, and therefore is also used in small and medium size engines. In addition, the
generally higher efficiencies exhibited by diesel engines makes diesel an ideal fuel for
large engines where operating costs must be minimized. Diesel is thus the most versatile
fuel for stationary reciprocating engines. Natural gas is used more than any other fuel for
large stationary reciprocating or turbine IC engines, typically operating pumps or
compressors on gas pipelines. Other fuels are also burned in stationary IC engines, but
their use is limited. Some engines burn heavy fuel oils, and a few burn almost any other
liquid fuel. Gaseous fuels such as sewer gas are sometimes used at wastewater treatment
plants where the gas is available. Stationary IC engines can be modified to burn almost
any liquid or gaseous fuel if the engine is properly designed and adjusted.
There are two methods for igniting the fuel in an IC engine. In spark ignition (SI), a spark
is introduced into the cylinder (from a spark plug) at the end of the compression stroke.
Fast-burning fuels, like gasoline and natural gas, are commonly used in SI engines. In
compression ignition (CI), the fuel-air mixture spontaneously ignites when the
compression raises it to a sufficiently high temperature. Compression ignition works best
with slow-burning fuels, like diesel. Larger engines may last for 20 to 30 years while
smaller engines (<1 MW) tend to have shorter life spans.
Stationary IC engines have efficiencies (total output/total input) that range from 25
percent to 45 percent. In general, diesel engines are more efficient than natural gas
engines because they operate at higher compression ratios. For future models, engine
manufacturers are targeting lower fuel consumption and shaft efficiencies up to 50-55
percent in large engines (>1 MW) by 2010. Efficiencies of natural gas engines, in
particular, are expected to improve and approach those of diesel engines.
D-2
Appendix E
ESI International Case Study Questionnaire
ESI International, Inc.
Suite 1100, 1660 L Street, NW
Washington, DC 20036
Voice: (202) 296-4797
Fax: (202) 331-1388
ESI International, Inc.
Case Study Questionnaire
Instructions: This questionnaire can be submitted electronically
to [email protected] or filled out by hand and faxed to (202)
331-1388. Electronic submissions are preferred. If you are
preparing an electronic submission, feel free to insert text or
data directly into the questionnaire. For assistance, please
contact Bill Gillespie at (202) 775-8868.
1. Contact Information
a. Person Preparing the Questionnaire
Name: ______________________________________________________
Address: ____________________________________________________
Telephone number: ___________________________________________
E-mail address: _____________________________________________
b. Engine Owner (If different from item 1a.)
Name: _______________________________________________________
Address: ____________________________________________________
Telephone number: ___________________________________________
E-mail address: _____________________________________________
c. Engine Operator (If different from item 1a.)
Name: _______________________________________________________
Address: ____________________________________________________
Telephone number: ___________________________________________
E-mail address: _____________________________________________
E-1
2. Facility Information
a. Describe the business where the engine is installed.
Please attach electronically (or in hardcopy) several paragraphs
that generally describe the nature of the business where the
engine is installed. This is an opportunity for you to tell us
about your company. Please send us company brochures or other
information about the firm if they are available.
b. Date engine was purchased: ___________________________________
3. Engine Specifications and Information
a. Make: ________________________________________________________
b. Model number: ________________________________________________
c. Serial number: _______________________________________________
d. Displacement: ________________________________________________
e. Operating horsepower: ____________ horsepower at _____________
revolutions per minute (rpm).
f. Turbocharged or naturally aspirated: _________________________
g. Fuel Consumption (gallons per hour; grams per brake horsepower
hour; or other unit of measure): _____________ per ______________
h. Engine load factor: __________________________________________
4. Fuel Specifications and Information
a. Type of Fuel:_________________________________________________
b. Fuel sulfur content (specify percent sulfur by weight or parts
per million (ppm)): _____________________________________________
c. Fuel heat content (Btu per gallon for diesel fuel, Btu per
cubic foot for natural gas):
_________________________________________________________________
d. Attach a fuel analysis report if available.
E-2
5. Engine Operating Information
a. Describe the engine’s principal function (for example, to
provide emergency electrical power, to pump water, to operate a
ski lift, etc.). Please attach a paragraph electronically or in
hardcopy.
_________________________________________________________________
b. How many hours do you operate the engine per year? ___________
hours per year.
c. Fuel Consumption (gallons per year for diesel fuel; therms or
cubic feet per year for natural gas):
Year
Fuel
Consumption
(gallons or
therms/year)
1
2
3
4
5
Alternative Method for Calculating Fuel Consumption:
If the number of hours the engine operated per year is known,
report the total number of hours and a fuel consumption rate
(gallons per hour or therms/cubic feet per hour).
Year
Operating
Hours
(hours/year)
1
2
3
4
5
Fuel consumption rate (gallons per hour or therms/cubic feet per
hour): _________________________________________________________
d. If the engine is used to generate electricity, provide
generation per year (kilowatt hours per year (kWhr/yr)):
Year
Electricity
Generation
(kWhr/yr)
1
2
3
4
5
e. If the engine provides power to drive a generator, does the
engine operate as an emergency back-up unit, a peaking unit, a
peak shaving unit, or other unit?
_________________________________________________________________
E-3
f. If the engine provides mechanical energy, please quantify, if
possible the mechanical work done (for example, gallons of water
pumped, etc.)
_________________________________________________________________
g. Describe engine operating problems if any:
_________________________________________________________________
6. Emission Control System Information and Specifications
a. Describe the installed emission control system:
_________________________________________________________________
Note: the emission control system may include the following
devices: diesel oxidation catalyst (DOC), diesel particulate
filter (DPF), exhaust gas recirculation (EGR), selective
catalytic reduction (SCR), or some other device. Please describe
the types of control equipment installed.
b. Why was the emission control system installed? (For example,
to meet state or federal permit requirements, etc.)
_________________________________________________________________
c. For each emission control system installed, provide:
Date installed: ___________________________________________
Make: _____________________________________________________
Model: ____________________________________________________
Serial number: ____________________________________________
Reagent (for SCR systems for example): ____________________
___________________________________________________________
7. Emission Control System Costs
Note: If you know the total installed cost of each emissions
control system, please provide that total cost here ____________
If you can provide disaggregated costs for each emission control
system, please complete items a through f below.
a. Purchased equipment costs:
________________________________________________________________
E-4
Note: Include the cost of emission control equipment, ancillary
equipment, instrumentation, etc.)
b. Sales taxes paid:
_________________________________________________________________
c. Freight paid:
_________________________________________________________________
d. Direct installation costs:
_________________________________________________________________
For installation of the emission control system, include direct
costs such as the costs of foundations and supports; equipment
handling and erection; providing electrical service, piping,
insulation, painting, etc. Indicate if any of these costs were
included in the purchased equipment costs, Item 7a above.
e. Indirect installation costs:
_________________________________________________________________
Note: For installation of the emission control system, include
indirect installation costs such as the costs of engineering,
construction and field expenses, contractor fees, start-up tests,
performance tests, studies and training. Indicate if any of
these costs were included in the purchased equipment cost, Item
7a, or direct installation costs, Item 7d above.
f. Contingency Costs:
_________________________________________________________________
Note: Include costs for equipment redesign and modifications,
cost escalations, delays in start-up, etc.
8. Emission Control System Operating Information
a. Describe the operation of the emission control system:
_________________________________________________________________
b. Engine Emission Rates (before installation of the emission
control system):
E-5
Emission Rate
Pollutant
Carbon Monoxide CO)
g/bhp-hr
lb/MMBtu
lbs/kWh
g/gal
ppm
Hydrocarbons (HC)
Particulate Matter
(PM)
Nitrogen Oxides
(NOx)
Reporting units:
g/bhp-hr = grams per brake horsepower hour
lbs/MMBtu = pounds per million British thermal units of heat
input lbs/kWh = pounds per kilowatt hour of electricity generated
g/gal = grams per gallon of fuel consumed
ppm = parts per million
At a minimum, please provide emission rate information in either
g/bhp-hr or lb/MMBtu. Provide emissions in lbs/kWh if available.
If you report emissions in units other than those shown above,
please define the units you use.
c. Test method used to determine engine emission rates:__________
_________________________________________________________________
d. Attach engine emission test data if available.
e. Engine emission rates after the emission control system was
installed:
Emission Rate
Pollutant
Carbon Monoxide CO)
g/bhp-hr
lb/MMBtu
lbs/kWh
g/gal
ppm
Hydrocarbons (HC)
Particulate Matter
(PM)
Nitrogen Oxides
(NOx)
Note: At a minimum, please provide emission rate information in
either g/bhp-hr or lb/MMBtu. Provide emissions in lbs/kWhr if
available. If you report emissions in units other than those
shown above, please define the units you use.
E-6
f. Percent reduction of air pollutants:
Pollutant
Carbon Monoxide (CO)
Percent Reduction
Hydrocarbons (HC)
Particulate Matter (PM)
Nitrogen Oxides (NOx)
g. Attach or provide the test method used to determine exhaust
pipe emission rates.
_________________________________________________________________
h. If a source testing company or state air quality agency tested
the engine after the emission control system was installed,
please provide or attach the emission test report.
_________________________________________________________________
i. Describe any operating problems associated with the emission
control system.
_________________________________________________________________
9. Emission Control System Operating Costs
a. Provide cost information for the operation of the emission
control system.
Labor:
Dollars per hour: _______________________________________________
Hours per year: _________________________________________________
Total labor costs per year: _____________________________________
Materials:
Describe the materials needed (for example, percent urea for SCR
reagents, etc.): ________________________________________________
_________________________________________________________________
Cost of reagent (dollars per gallon): ___________________________
E-7
Reagent consumption rate (gallons per hour): ____________________
Reagent use (hours per year): ___________________________________
Reagent costs per year: _________________________________________
Electricity cost per kilowatt-hour: _____________________________
Electricity hours per year: _____________________________________
Electricity costs per year: _____________________________________
Other material costs per year: __________________________________
Total material costs per year: __________________________________
10. Emission Control System Maintenance
a. Describe the maintenance requirements of the emission control
system:
_________________________________________________________________
b. How often is maintenance of the emission control system
performed:
_________________________________________________________________
11. Emission Control System Maintenance Costs
Labor:
Dollars per hour: _______________________________________________
Hours per year: _________________________________________________
Total labor costs per year: _____________________________________
Materials:
Describe the materials needed: __________________________________
_________________________________________________________________
Expected life of catalyst (years): ______________________________
Cost of catalyst replacement: ___________________________________
Total material costs per year: __________________________________
E-8
Thank you for participating in this case study.
Please send the completed questionnaire to:
[email protected]
or
Bill Gillespie
ESI International, Inc.
1660 L Street, NW, Suite 1100
Washington, DC 20036-5603
Telephone: (202) 775-8868
Fax: (202) 331-1388
E-9